by fabrizio ferri* march 30, 2014 *associate professor ... · tool (favored by a benign accounting...
TRANSCRIPT
Say on Pay
By Fabrizio Ferri
March 30 2014
Associate Professor Columbia Business School Columbia University
This chapter is copyright 2012 Edward Elgar Used by permission of Edward Elgar for personal use only not for resale or reuse
Introduction
On July 21 2010 President Obama signed into law the Dodd-Frank Wall Street Reform and
Consumer Protection Act (hereafter Dodd-Frank Act)1 One provision of the Act mandates that
beginning with annual meetings on or after January 21 2011 US publicly traded firms allow
shareholders a nonbinding vote on executive pay known as ldquosay on payrdquo (hereafter SOP)2
The purpose of this chapter is to present key insights from the academic research on the
economic consequences of SOP The chapter is organized around five parts Section I provides a
brief history of SOP highlighting its role in the broader context of a gradual shift toward greater
shareholder empowerment The rich set of legislative events accompanying the adoption of SOP
in many countries has provided fertile ground for empirical studies Section II and Section III
review the evidence in such studies focusing on the effect of SOP respectively on executive
pay and firm value followed by some concluding remarks in Section IV3
1 Dodd-Frank Wall Street Reform and Consumer Protection Act Pub L No 111- 203 sect951 124 Stat 1376 1899
(2010) (ldquoDodd-Frank Section 951Prime) Dodd-Frank Section 951 amended the Securities Exchange Act of 1934 by
adding Section 14A (codified as amended at 15 USC sect78n-1) (ldquoExchange Act Section 14Ardquo)
2 Dodd-Frank Section 951 also mandates a non-binding vote on the frequency of future say on pay votes (known as
say-when-on-pay vote) with a choice between an annual a biennial or a triennial frequency This vote on the
frequency of future SOP votes must be held every six years At most firms shareholders overwhelmingly favored an
annual frequency For an analysis of determinants and consequences of say-when-on-pay votes see Ferri and Oesch
(2013)
3 Some of the studies on SOP also examine the determinants of SOP votes Since by and large these studies show
that the key determinants are the same as for other types of shareholder votes (eg proxy advisorsrsquo
recommendations ownership composition performance) I devote little time to this topic and refer the reader to the
Chapter on the Theory and Practice of Corporate Voting by Paul H Edelman and Randall S Thomas
A caveat the academic research on SOP is growing rapidly as more data become available over
time and across countries As a result many studies cited here are in the form of working papers
and their findings should be viewed as preliminary Also I may have missed some of the most
recent studies Finally I apologize if I do not discuss in equal depth all the studies and tend to
focus instead on the work I am more familiar with including my own
I A Brief History of Say on Pay
While CEO pay has made headlines and captured the attention of politicians and policy makers
for many decades (Murphy 2012) it has arguably received even greater scrutiny over the last two
decades During the New Economy of the 1990s the growing use of stock options as incentive
tool (favored by a benign accounting treatment) led to a rapid increase in CEO pay with the
average CEO-to-worker pay ratio peaking at more than 400 in 2000 up from 18 in 1965 (Mishel
and Sabadish 2012) As the dot-com bubble began to burst and a series of accounting and
governance scandals unfolded (eg Worldcom Enron) stock options and executive pay were
blamed for providing perverse incentives to manipulate financial reports and the stock price At
the same time these governance scandals led many institutional investors to take a more active
role in monitoring corporations spurring a wave of shareholder activism A whole new industry
of governance experts and intermediaries (governance ratings agencies proxy advisors)
emerged while influential academic studies documented a large impact of governance quality on
firm value (Gompers Ishii and Metrick 2003 Bebchuk Cohen and Ferrell 2009) Calls for
policy reforms empowering shareholders became louder and corporate governance took center
stage in the policy debate (Bebchuk 2005) Like other institutional investors union pension
funds began to take a more active role Because union pension funds tend to be well diversified
holding very small stakes in thousands of firms they could not exert their influence by buying
large stakes in firms Hence they resorted to lsquolow-costrsquo tools of activism (Ferri 2012) such as
shareholder proposals and shareholder votes on uncontested director elections Tellingly the
percentage of shareholder proposals filed by union pension funds increased from 136 in 1997
to 436 in 2003 when union pension funds surpassed individual investors in terms of number
of proposals and continued to grow thereafter (Ertimur Ferri and Stubben 2010) Not
surprisingly executive pay became a central focus of unionsrsquo governance activism Between
2003 and 2010 led by union pension fundsrsquo efforts the frequency of and voting support for
compensation-related shareholder proposals and compensation-related lsquovote-norsquo campaigns
against directors up for election increased quickly For example Ertimur Ferri and Muslu (2011)
report that among SampP 1500 firms there were approximately 66 proposals per year in the 1997-
2002 period (averaging 162 votes in favor) compared to about 160 proposals per year in the
2003-2007 period (289 votes in favor) New types of shareholder proposals emerged (eg
proposals to introduce performance-based vesting conditions in equity grants expense stock
options subject severance payments to shareholder approval) gaining higher voting support
While these votes were nonbinding firms began to respond particularly when votes were
withheld from compensation committee members up for re-election (Del Guercio Seery and
Woidtke 2008 Cai Garner and Walkling 2009 Ertimur Ferri and Muslu 2011)
Meanwhile an important development had taken place in the United Kingdom (UK) where in
2002 the UK government had introduced a mandatory annual advisory vote on executive
compensation known as ldquosay on payrdquo largely in response to a perceived increase in US-style
lsquofat catrsquo executive pay packages among UK firms During the first say on pay proxy season a
failed say on pay vote at Glaxo SmithKline (with 507 of the votes cast against approval of the
remuneration report) made headlines around the world Shareholders had objected to an
estimated pound22 million severance arrangement for the CEO (reflecting a two-year notice period)
lack of challenging performance targets and the presence of a retesting provision in the stock
option plan4 Glaxo SmithKlinersquos board responded to the vote by launching an extensive
consultation process with shareholders and adjusting the compensation package accordingly
(Ferri and Maber 2013)
In the United States the American Federation of State County and Municipal Employees a
union pension fund took notice of the UK experience and rallied other activist investors to
submit shareholder proposals requesting the adoption of SOP (Ferri and Weber 2009) Between
2006 and 2010 SOP proposals were submitted and voted upon at hundreds of US firms
averaging more than 43 votes in favor and often winning a majority vote with increasing
success over the years (Cuntildeat Gine and Guadalupe 2013) High-profile executive pay scandals
(eg option backdating the large severance package awarded to Nardelli at Home Depot) helped
activists in making their case for more ldquosay on payrdquo among investors and policy makers While
submitted at individual companies the objective of the SOP proposals was to induce widespread
adoption of SOP either via voluntary adoption (which turned out to be rare) a listing standard
or regulatory intervention Policy makers took notice and on April 20 2007 the House of
Representatives passed a bill requiring a nonbinding annual shareholder vote on executive
4 Retesting provisions in the performance-based vesting conditions of equity grants allow to reevaluate in
subsequent years performance targets not achieved during the initial measurement period rather than allowing the
equity grant to lapse As such they were criticized in the UK as a form of lsquoreward for failurersquo
compensation (hereafter SOP Bill)5 On the same day then-Senator Barack Obama introduced a
companion bill in the Senate (S1181)
But it was the financial crisis of 2007-2008 to accelerate the path toward mandatory
adoption of SOP Growing income inequality public outrage over Wall Street excesses (eg the
large bonuses paid at AIG) and banksrsquo bailouts and the perception that executive pay played a
role in inducing excessive risk-taking pressured policy-makers to take action with respect to
executive pay Many of the legislative proposals discussed in the House and Senate in 2008 and
2009 contained a SOP provision (see Table 1 in Larcker Ormazabal and Taylor 2011) Also
holding a SOP vote was a mandatory condition for firms to receive funds under the Troubled
Asset Relief Program (TARP) During the 2008 campaign both Presidential candidates
expressed support for SOP As these events unfolded the debate on merits and drawbacks of
SOP heated up Critics of SOP argued that at best SOP votes would be ignored (because
nonbinding) and at worst would cause directors to pander to shareholders with special interests
or lacking the required expertise and sophistication ultimately resulting in the adoption of
suboptimal pay practices (Kaplan 2007 Bainbridge 2008)6 They also cautioned that the high
5 House Bill 1257 Shareholder Vote on Executive Compensation Act
6 Analytical studies have also raised questions on the informativeness of SOP votes Ozbas and Matsusaka (2013)
note that SOP votes only reveal shareholder preferences for the current compensation plan compared to the
alternative plan that will be adopted if the current plan is rejected but such alternative plan is unclear in practice
casting doubts on whether the shareholder votes on SOP may be truly informative Levit and Malenko (2011) further
suggest that shareholder votes may not be informative about shareholder preferences when shareholders ignore their
own information and vote strategically conditioning on the chance of casting a pivotal vote
cost of analyzing executive pay at thousands of firms would lead shareholders to outsource
voting decisions to proxy advisors who in turn would minimize their own costs by promoting
one-size-fits-all compensation practices which would hurt firm value Finally SOP opponents
noted that shareholders had already tools to express their views on compensation matters such as
the ability to submit shareholder proposals on compensation and withhold votes from directors
responsible for compensation packages Supporters of SOP argued that enhanced shareholder
voice (as formalized in a SOP vote) and reputation concerns would help boards overcome
psychological barriers to negotiating with CEOs on behalf of shareholders resulting in more
efficient compensation contracts and more dialogue between boards and shareholders (Bebchuk
2007) They also pointed out to growing evidence of shareholdersrsquo sophistication in casting
informed votes Finally they argued that SOP would be more effective than shareholder
proposals (limited to a single issue) and less disruptive than a confrontational vote against an
otherwise valuable director More fundamentally critics and supporters of SOP disagreed on
whether existing compensation contracts were the result of an efficient labor market for talent or
instead the expression of management power over captive boards
Finally in 2010 a SOP provision found its way in the Dodd-Frank Act Ironically SOP a
provision designed to provide more alignment between shareholder and management was
adopted as part of a financial reform package aimed at deterring lsquoexcessiversquo risk taking
(excessive from the point of view of social welfare) which may be the result of the alignment of
interests between shareholders and management
As of today non-binding SOP votes (on the compensation report) have been mandated in
Australia Belgium Denmark Israel Italy Portugal while Netherlands Sweden Norway South
Africa and Switzerland adopted some version of a binding SOP vote (on the future compensation
policy) which is now also being considered by the European Union7
When putting SOP into its historical context it is also important to appreciate that SOP has come
to epitomize a broader movement toward greater shareholder democracy beyond the issue of
executive pay For example during the same period shareholder activists have long
(unsuccessfully) lobbied for a proxy access rule that would have made it easier for shareholders
to oust corporate directors and nominate their own candidates Hence many observers often
view the successes and failures of SOP as speaking to the potential effects of other reforms
aimed at increasing shareholder power making the academic research on SOP all more relevant
II The Effect of Say on Pay on Executive Compensation
A Effect of SOP votes on level and composition of executive pay
As discussed earlier United Kingdom was the first country to adopt SOP in 2002 for publicly
traded firms Ferri and Maber (2013) report that in most cases shareholders voted in favor of
compensation plans Failed SOP votes (ie greater than 50 of votes against) were rare (2 of
the sample) though highly publicized in the press However about one-fourth of the sample
firms received more than 20 of votes against
To capture the overall effect of SOP on CEO pay Ferri and Maber (2013) examine the
sensitivity of CEO pay to its economic determinants over the 2000-2005 period (ie before and
after the introduction of SOP regulation) and document a significant increase in the sensitivity of
7 Some countries have already modified their first SOP legislation In 2013 the UK added to the non-binding
backward-looking SOP vote on the current compensation report a binding forward-looking vote on the proposed
compensation policy In 2012 Australia modified the non-binding SOP vote to introduce the so-called ldquotwo-strikesrdquo
rule Under the rule if 25 of shareholders vote against a companyrsquos remuneration report at two consecutive annual
general meetings the entire board may have to stand for re-election within three months For more details about the
international development of SOP see Thomas and Van der Elst (2013)
CEO pay to poor performance They also find that this increase does not occur for a subset of
UK firms exempted from the SOP regulation (firms traded on the Alternative Investment Market
a sub-market of the London Stock Exchange with a more flexible regulatory system) suggesting
that the result reflects the impact of SOP rather than a general trend affecting all firms Besides
Ferri and Maber (2013) find that the increase is more pronounced in firms experiencing high
voting dissent and firms with high abnormal CEO pay before the adoption of SOP consistent
with a causal impact of SOP In contrast they fail to find any effect of SOP on the level of pay
Correa and Lel (2013) examine the effect of SOP laws on CEO pay using a large cross-country
sample of about 103000 firm-year observations from 39 countries including 12 countries that
adopted some (advisory or binding) version of SOP They essentially compare post-SOP CEO
pay at firms of countries adopting SOP to a control sample which includes all non-SOP
observations (that is pre-SOP CEO pay in countries eventually adopting SOP as well as CEO
pay in countries never adopting SOP) They find that firms in a post-SOP regime (i) exhibit
lower CEO pay levels (but only in countries adopting an advisory SOP vote) a finding driven by
a relative decline in equity awards and limited to CEOs and not the other top executives (ii)
higher pay-performance sensitivity (the authors do not look at positive and negative performance
separately) The current version of the paper does not examine whether these findings are more
pronounced in (or driven by) firms with excess CEO pay before the adoption of SOP and in
firms experiencing adverse SOP votes
The drawback of the research design in Correa and Lel (2013) is that it is not clear whether the
results reflect time-series changes within countries adopting SOP (pre vs post) or difference
between firms in SOP and non-SOP countries For example further tests in the paper suggest
that the lower CEO pay levels for post-SOP observations are not driven by a pay decrease
relative to the pre-SOP period (in fact country-by-country analyses indicate no change in CEO
pay levels after the adoption of SOP) but by the comparison with non-SOP countries Also while
the authors employ reasonable econometric solutions to deal with this issue some concerns
remains as to whether non-SOP countries can be a good benchmark for what would have
happened in SOP countries had SOP not been adopted (in other words if only countries with
problematic CEO practices adopt SOP it is possible that changes in CEO pay reflect factors
leading to the adoption of SOP rather than the effect of SOP laws per se) Notwithstanding these
concerns these findings are intriguing and future versions of this recent study have the potential
to provide more conclusive evidence on the effect of SOP
While in the US SOP was mandated by the Dodd-Frank Act its implementation was left to the
Securities and Exchange Commission (SEC) The final rule issued in January 2011 exempted
lsquosmallrsquo firms (ie firms with a public float below $75 million) from the SOP provision for two
years Iliev and Vitanova (2013) exploit this setting to estimate the effect of SOP on CEO
compensation In particularly using a regression discontinuity design they compare changes in
CEO pay of firms just above the SEC-imposed threshold to changes in CEO pay of firms just
below the threshold (but otherwise quite similar) They find no differences in terms of level and
composition of CEO pay and in terms of golden parachutes concluding that SOP had no impact
on CEO pay While the identification strategy chosen by the authors has the benefit of reducing
endogeneity issues it comes with the price of focusing only on fairly small firms (those around
the $75 million public float threshold) where CEO pay may not be a problem in the first place
(in which case the lack of impact of SOP would not be surprising but would not speak to its
impact at large firms traditionally the target of compensation-related criticism)
Cuntildeat Gine and Guadalupe (2013) also use a regression discontinuity design but in a different
setting In particular they analyze nonbinding shareholder proposals to adopt SOP (250
proposals between 2006 and 2010) Comparing the changes in CEO pay for firms voluntarily
adopting SOP in response to the proposals to firms not adopting SOP would be difficult since
the adoption decision is endogenous Similarly it would be difficult to compare changes in CEO
pay between firms where the proposal is passed (and thus significantly more likely to be
adopted) and firms where the proposal fails to be approved since the voting outcome is
endogenous as well
Hence Cuntildeat Gine and Guadalupe (2013) use instead a regression discontinuity design that
compares firms where the proposals receive slightly more than 50 of the votes (the threshold
for approval) to otherwise similar firms where the proposals receive slightly less than 50
Using this approach they conclude that the passing and adoption of SOP proposals is not
associated with changes in the level of CEO pay nor its composition and structure (mix of cash
and equity equity incentives level of perks and deferred pay) In some sense this finding
complements the one in Iliev and Vitanova (2013) in that it is based on large firms (shareholder
proposals are typically submitted at large SampP 500 firms) However both Cai and Walkling
(2011) and Cuntildeat Gine and Guadalupe (2013) show that firms targeted by SOP proposals do not
exhibit excess CEO pay or weaker governance8 Hence the lack of an effect at these firms may
8 Activists submitting SOP proposals chose to focus on a broad sample of large firms rather than target only firms
with problematic CEO pay practices to obtain visibility and show to policy-makers that investorrsquos support for SOP
was not limited to problematic firms (Ferri and Weber 2009)
not be surprising More generally the findings cannot be necessarily generalized to firms away
from the 50 threshold or to firms not targeted by SOP shareholder proposals
B Effect of SOP votes on compensation practices
Analyses of the effect of SOP on level and composition of CEO pay may not capture a number
of other changes to compensation contracts that may be induced by SOP votes but are not
reflected in measure of CEO pay For example the introduction of performance-based vesting
provisions is typically not reflected in the estimates of fair value of equity grants used by
researchers in measuring CEO pay Many other changes (eg terms of severance packages) will
only be reflected in measures of CEO pay contingent upon the occurrence of certain events
Hence it is important to complement the evidence in section IIA with an analysis of firmsrsquo
disclosures of changes to observable provisions of compensation contracts Another benefit of
this approach is that it captures the changes that boards explicitly present as a response to SOP
votes and thus it acts as a reality check on regression-based inferences For example a
regression-based finding of say a decrease in CEO pay level would be more credible if
supported by evidence that firms disclose a decision to reduce target levels of CEO pay in
response to an adverse SOP vote (an action that presumably they have an incentive to disclose to
get rewarded by shareholders in subsequent votes)
Two studies collect data on specific changes made to compensation contracts explicitly in
response to SOP votes in the UK and the US using firmsrsquo disclosures in their proxy statements
Ferri and Maber (2013) examine the compensation reports of a sample of UK firms before and
after the first SOP vote and find that firms experiencing higher voting dissent were significantly
more likely to remove compensation provisions criticized by investors as lsquorewards for failurersquo
relative to a matched sample of firms experiencing lower dissent For example they report that
among firms with long notice periods (implying larger severance payments ) the percentage of
high dissent firms that shortened them after the vote (800) was significantly higher than before
the vote (200) and also significantly higher than among low dissent firms after the vote
(333) They report similar findings for another controversial practice namely the presence of
retesting provisions in the performance-based vesting conditions of equity grants Most firms
indicate that these changes were the result of consultations with their major institutional
investors Firmsrsquo responsiveness to SOP votes was lsquorewardedrsquo with a substantial increase in
favorable SOP votes at the subsequent annual meeting Ferri and Maber (2013) also find that
many firms experiencing low voting dissent at the first SOP vote had removed these practices
before the vote9 highlighting the importance of accounting for ex ante effects when assessing the
impact of a new regulation
Ertimur Ferri and Oesch (2013) examine the effect of SOP on compensation practices in the US
in 2011 (first proxy season under mandatory SOP) Their key findings are generally similar to
the evidence from the UK (i) failed SOP votes are rare (about 2 of the sample) though cases
of substantial dissent are more frequent (ii) more than 55 of the firms experiencing significant
voting dissent respond by making material changes to their compensation plan during the
subsequent year (eg introduction of performance-based vesting conditions in equity grants use
of tougher performance targets removal of perks and tax gross-ups removal of controversial
provisions from severance contracts etc) usually in consultation with major institutional
9 In their sample 700 of low dissent firms shortened their notice periods during the year before the SOP vote
investors (iii) firms making changes to their compensation plans receive greater voting support
at the following SOP vote
Ertimur Ferri and Oesch (2013) also highlight the strong influence of the recommendations
released by proxy advisors (particularly Institutional Shareholder Services ISS) with a negative
recommendation being associated with 25 more votes against the compensation plan Further
evidence of the significant influence of ISS is the striking discontinuity in the relation between
the extent of SOP voting dissent and firmsrsquo responsiveness to the vote the percentage of firms
making compensation changes in response to SOP votes jumps from 32 to 72 around the
30 voting dissent threshold Why After the 2011 proxy season ISS had indicated that firms
failing to ldquoadequatelyrdquo respond to SOP voting dissent above 30 would receive a negative
recommendation in 2012 on the SOP proposal and on the election of compensation committee
members
Finally similar to the UK experience there is also anecdotal evidence of analogous
compensation changes being made by US firms ahead of the SOP vote to avoid a negative
proxy advisor recommendation and an adverse shareholder vote (Thomas Palmiter and Cotter
2013 Larcker McCall and Ormazabal 2013)
III The Effect of Say on Pay on Firm Value
In this section we examine the empirical studies on the overall effect of SOP on firm value
A Event studies around the adoption of Say on Pay regulations
As mentioned in the Introduction on April 20 2007 the House of Representatives passed a SOP
Bill by a 2-1 margin On the same day then-Senator Barack Obama introduced a companion
bill (S1181) in the Senate (which was then put on hold by the Senate Banking Committee)
While the SOP Billrsquos approval was expected (Democrats were in control of the House and
supported the SOP Bill) the 2-1 margin was unexpected suggesting some support for SOP
among Republicans
Cai and Walkling (2011) examine the market reaction to the Housersquos approval of the SOP Bill
for a sample of firms in the SampP 1500 index and document a positive reaction for firms more
likely to benefit from greater shareholder voice over executive pay namely firms with high
abnormal CEO cash pay (defined as the difference between actual CEO cash pay and the level
predicted based on known economic determinants such a size performance industry) firms
with low pay-for-performance sensitivity firms with a history of shareholders willing to vote
against compensation-related management proposals and firms with a history of responsiveness
to shareholder pressure on compensation issues In other words they find a positive reaction in
firms where the compensation problems are more severe and a say on pay vote is more likely to
trigger a response (eg more shareholders willing to vote against management and greater firmsrsquo
propensity to respond to an adverse vote) Note however that they do not find a significant
impact for firms with high abnormal equity and total CEO pay
Larcker Ormazabal and Taylor (2011) also examine the market reaction to the Housersquos approval
of the SOP Bill (in addition to other regulatory events related to executive pay and other
governance provisions) Similar to Cai and Wakling (2011) they fail to find a significant impact
for firms with high abnormal total CEO pay (they do not examine the price reaction for firms
with abnormal cash CEO pay or firms with lower pay-performance sensitivity)
A concern with both studies is that the Housersquos approval of a bill does not guarantee passage in
the Senate and approval by the White House In fact the press at the time reported that the
prospects of the bill in the Senate were uncertain (New York Times 2007) and the Bush White
House openly opposed the Bill (Associated Press 2007) Hence it is not clear the extent to
which this event increased the likelihood of a say on pay legislation Arguably as noted in
Section I the only lsquoeventrsquo that substantially increased the likelihood of say on pay legislation
was the financial crisis and the related perception of compensation abuses (eg outrage over AIG
bonuses)
Ferri and Maber (2013) argue that the announcement of the submission of SOP regulation to the
Parliament in the United Kingdom in June 2002 offers a more powerful setting for an event
study since it was largely unexpected and increased substantially the probability of SOP
adoption (Parliamentrsquos approval was virtually guaranteed) Using this event they document
positive abnormal returns for firms with excess CEO pay combined with poor performance and
for firms with controversial pay practices that weaken the penalties for poor performance (those
practices were often removed in response to SOP votes as discussed earlier in Section IIB)
They interpret their findings as consistent with shareholders perceiving SOP as a value
enhancing monitoring mechanism for firms with low pay-to-poor-performance sensitivity
Finally a recent study by Iliev and Vitanova (2013) examines the market reaction around the
announcement of the SEC final SOP rule that on January 25 2011 exempted small firms (ie
firms with a public float below $75 million) from adopting SOP for two years (press reports at
the time noted that the exemption could become permanent) An appealing feature of this setting
is that the SEC decision was an unexpected reversal from the rule proposed in October 2010 rule
under which no publicly traded firm would be exempted
Iliev and Vitanova (2013) compare the market reaction to this announcement for firms just above
the SEC-imposed threshold and thus subject to the SOP rule and firms just below the threshold
(but otherwise quite similar) and thus exempted from SOP While they find not significant
abnormal returns for either group the abnormal returns for exempted firms are significantly
more negative a finding that the authors interpret as evidence of a positive value effect of
mandatory SOP While the setting is an interesting one some concerns remain as exempted
firms are fairly small and the magnitude of the executive pay problem (if any) is unlikely to be
large enough to explain the differential market reaction Also the same study finds no effect on
CEO pay level and composition (see Section IIA) Hence the reason for the differential stock
price reaction remains unclear It would be helpful to know if the event study result is driven by
(or stronger for) exempted firms with excessive pay or problematic pay practices
B Event studies around Shareholder Proposals to Adopt Say on Pay
Starting in 2006 shareholder activists led by union pension funds submitted shareholder
proposal to adopt SOP at hundreds of US firms in an attempt to induce voluntary or mandatory
widespread adoption of SOP
Cai and Walkling (2011) examine the market reaction to proxy filings and annual meetings of
113 firms targeted by nonbinding shareholder proposals to adopt SOP between 2006 and 2008
On average they find insignificant returns around both the submission of the proposal (proxy
filing date) and the vote on the proposal (annual meeting date) But they also find that when the
proposals are filed by union pension funds the abnormal returns (still insignificant) are more
negative than when filed by other activists and that when the proposal is defeated the abnormal
returns (while insignificant) are more positive than when the proposal is passed (it is not clear
whether this result is driven by union-sponsored proposals) They interpret these findings as
evidence that the market views SOP proposals filed by union pension funds are driven by special
interests rather than value maximization10
However caution is required in interpreting this
evidence Virtually all activists filing SOP proposals were coordinated by a group of investors
(mostly but not only union pension funds) who made available to other activists a template for
SOP proposals and a list of target firms (Ferri and Weber 2009) Hence the distinction between
union and non-union proponents in the context of SOP proposals may be only apparent More
importantly the list of target firms was publicly available months before the proxy filing dates
so it is not clear whether the proxy statements contained any new information Similarly the
voting outcome of the SOP proposals may largely be anticipated (based on the composition of
institutional owners proxy advisorsrsquo recommendations etc) Finally the analyses do not control
for other (potentially new) information contained in proxy statements (eg executive
compensation report other items up for a vote at the annual meeting) or other events occurring at
the annual meeting (eg shareholder votes on other items)
Cuntildeat Gine and Guadalupe (2013) also examine the market reaction to the voting outcome of
nonbinding shareholder proposals to adopt SOP but to alleviate these concerns they employ a
lsquofuzzyrsquo regression discontinuity design (RDD) essentially comparing the stock price reaction to
SOP proposals that pass by a small margin to the reaction to SOP proposals that fail by a small
10
Consistent with this interpretation the authors also find that SOP proposals generally targeted larger firms rather
than firms with excessive pay or lower pay-performance sensitivity However as noted by Ferri and Sandino (2009)
activists typically submit shareholder proposals aimed at promoting policy reforms at a broad sample of large firms
rather than the firms that would most benefit from the proposals because they believe that showing widespread
support for the proposal across all types of firms enhances its credibility with policy makers
margin (Cuntildeat Gine and Guadalupe 2012 Ertimur Ferri and Oesch forthcoming) The underlying idea
is that firms around the threshold are likely to have similar characteristics (indeed they do as the
authors show) but differ in the likelihood of implementation which is typically much higher for
proposals passing the threshold (Ertimur Ferri and Stubben 2010 Thomas and Cotter 2007)
Indeed the authors show that the likelihood of subsequently adopting SOP is 48 higher when
the SOP proposal passes by a small margin relatively to when it fails by a small margin Besides
because in these close-call situations the voting outcome is uncertain its resolution (pass or fail)
is likely to convey new information about the likelihood of SOP adoption making this setting
suitable to an event study11
Using this RDD approach Cuntildeat Gine and Guadalupe (2013) find that on the day of the vote a
SOP proposal that passes by a small margin yields an abnormal return of 24 relative to one
that fails (after controlling for other proposals voted upon at the same meeting) and estimate the
lsquofullrsquo value of SOP at 46 (after taking into account the increase in the probability of SOP
implementation)12
Aside from the issues discussed in Section IA (ie generalizability to other
firms) this estimate seems too large to reflect the present value of future reductions in excess
CEO pay (as noted by the authors their estimate of the value of SOP is equivalent to the
estimated value of removing two anti-takeover provision based on Cuntildeat Gine and Guadalupe
2012) particularly because (i) the authors find no evidence of subsequent changes in levels and
composition of CEO pay for firms implementing SOP (see Section IIA) an (ii) the sample firms
11
Consistent with this observation the authors find no market reaction to the voting outcome of SOP shareholder
proposals away from the threshold
12 Since the outcome of the vote is not binding the 24 market reaction only reflects the expected increase in the
probability of SOP adoption and thus understates the value of the SOP provision
do not appear to be characterized by excess CEO pay However the authors also find significant
improvements in subsequent operating performance and efficiency as a result of passing SOP
proposals and thus conjecture that the SOP vote is viewed as a tool to express a vote of
confidence in management performance more than a way to change compensation practices and
that the large positive market reaction reflects expected performance improvements under this
tighter monitoring regime While this is an interesting idea it remains unclear what additional
ldquoteethrdquo a SOP vote may provide over existing (non-compensation related) monitoring
mechanisms shareholders can express (and do often express) their dissatisfaction with
management performance when voting on director elections (Cai Garner and Walkling 2009)
C Event studies around compensation changes induced by SOP
Two studies examine announcements of changes to compensation plans related to SOP Larcker
McCall and Ormazabal (2013) examine the market reaction to (non-contaminated) compensation
changes disclosed in 8-K filings by Russell 3000 firms in the US during the year before the first
SOP vote They find that changes that appear to be made to avoid a negative proxy advisorrsquos
recommendation and thus a negative SOP vote are associated with a negative abnormal return
of -044 whereas other compensation changes are not associated with a significant stock price
reaction Ertimur Ferri and Oesch (2013) perform a similar analysis but focusing on
compensation changes made after the first SOP vote (and explicitly in response to the vote) by
firms receiving a negative recommendation and a high voting dissent in 2011 They fail to find
significant abnormal returns even for the subset of compensation changes that resulted in a
positive recommendation and low dissent in 2012 (and thus were presumably perceived to be
adequate and material by proxy advisors and voting shareholders)
D Other approaches effect of SOP on Tobinrsquos Q
In their cross-country study Correa and Lel (2013) also examine the effect of SOP laws on
Tobinrsquos Q and report a 36 increase in firm value following the adoption of the SOP laws
(more precisely a 36 higher firm value for post-SOP observations relative to a control sample
that pools together pre-SOP and non-SOP observations) They acknowledge that this increase in
firm value is too large to be justified by the relative decrease in CEO pay that they document and
suggest (but do not test) that it may reflect better alignment of pay and performance13
A key
concern is that higher valuation in countries with SOP laws may reflect other governance
changes introduced at the same time While the study controls for other compensation-related
laws there may be other non-compensation regulations introduced with SOP and with a
potentially larger impact on firm value
IV What have we learned
Adoption of SOP in the US has been long advocated by those who contend that executive pay is
a manifestation of rather than a solution to the agency problem and by those who favor great
shareholder involvement in corporate decisions A growing body of research is starting to
investigate the effect of SOP on executive pay and firm value It is premature to draw definitive
conclusions also because these studies differ in methodologies and settings but three points
seem to emerge First there is robust evidence that boards respond to SOP votes when
13
They also find that the firm value increase is higher when the relative decrease in CEO pay results in a lower pay
differential between CEO and other top managers (CEO pay slice) implying that a source of value creation may be
the reduced pay inequality among the top management team But this seems unlikely to explain the change in
Tobinrsquos Q since the increase occurs for both firms with advisory SOP laws and firms with binding SOP laws and
there is no change in CEO pay slice in firms subject to binding SOP laws
shareholders use it ldquovoicerdquo is heardrdquo Both in the US and UK studies show that not only firms
failing to win the SOP vote but also firms facing substantial dissent (20-30 of the votes
against) make changes to their compensation plans in consultation with institutional investors
and proxy advisors Second in both the US and UK institutional investors appear to use the
power of SOP votes to pressure firms into strengthening the link (or the perceived link) between
pay and performance (particularly on the downside) but not to pressure firms to reduce target
levels of pay suggesting a reluctance of institutional investors to lsquoregulatersquo executive pay Third
and consistent with the above most studies examining the aggregate effect of SOP on CEO pay
find some increase in pay-performance sensitivity but not effect on the level and growth of CEO
pay While some observers may interpret the lack of an effect on CEO pay levels as evidence of
the failure of SOP one should note that SOP per se is a neutral tool Its impact depends on what
investors choose to lsquosay on payrsquo Fourth based on my interpretation of the existing research to
date there is little evidence of an effect of SOP on firm value The studies documenting a
positive price impact appear subject to alternative interpretations or not plausible in their
findings (eg they fail to find an effect on CEO pay or the effect is too small to justify the
documented price impact hence it remains unclear what the source of value creation is) Also
there seems to be no stock price reaction to the SOP-induced compensation changes Finally
given the nature of these SOP-induced changes while some of them may be beneficial it seems
hard to believe that they will have a statistically detectable impact on firm value except perhaps
in a handful of firms where the compensation plan is subject to a complete overhaul
Perhaps one explanation for the weak impact of SOP is that executive pay problems were
overstated or that by the time SOP was introduced (more than a decade after the Enron-type
scandals that led to calls for greater shareholder voice) they had been already addressed via
other mechanisms (hedge fund activism monitoring by institutional investors vote-no
campaigns against compensation committee members SEC-mandated pay disclosures)
Overall though based on the evidence to date it does not appear that SOP had a major impact14
Was the adoption of SOP an lsquooptimalrsquo choice from a social welfare point of view This is a
difficult question to answer given the challenges of identifying and measuring all the potential
costs and benefits associated with SOP (likely to change over time and differ across countries)
Perhaps the most positive view of SOP is that its introduction prevented the adoption of other
more radical and intrusive regulatory measures (eg CEO pay caps) In that sense it may be
viewed as an lsquooptimalrsquo answer (ie the lesser of two evils) to the political pressure to reform
executive pay during the financial crisis
14
This interpretation of the evidence is also consistent with the prediction of analytical studies Ozbas and
Matsusaka (2013) model the benefits and costs of shareholder empowerment distinguishing between the right to
approve and the right to propose They show that permitting shareholders to propose directors or policies can cause
value-maximizing managers to take value-reducing actions to accommodate activist investors with non-value-
maximizing goals As for the right to approve it is weakly beneficial that is approval rights (such as a SOP vote)
are generally beneficial for shareholders in that they limit the managerrsquos ability to pursue private benefits at
shareholder expense but they have minimal impact on managerial actions and firm value (because the manager in
effect can threaten shareholders with an undesirable status quo if they do not approve the managerrsquos proposed
action)
References
Associated Press 2007 Administration opposes say on pay bill April 19
Bainbridge S 2008 Remarks on say on pay An unjustified incursion on director authority
UCLA School of Law Research Paper No 08-06
Bebchuk L 2005 The Case for Increasing Shareholder Power Harvard Law Review 118833-
917
Bebchuk L 2007 Written testimony submitted before the Committee on Financial Services
United States House of Representatives Hearing on Empowering Shareholders on Executive
Compensation March 8
Bebchuk LA A Cohen and A Ferrell 2009 What Matters in Corporate Governance Review
of Financial Studies 22783-827
Cai J J Garner and R Walkling 2009 Electing Directors Journal of Finance 642387-2419
Cai J and R Walkling 2011 Shareholdersrsquo Say on Pay Does it Create Value Journal of
Financial and Quantitative Analysis 46299ndash339
Coates IV JC (2009) Testimony before the Subcommittee on Securities Insurance and
Investment of the Committee on Banking Housing and Urban Affairs United States Senate
July 29 2009 httppapersssrncomsol3paperscfmabstract_id=1475355
Correa R and U Lel 2013 Say On Pay Laws Executive Compensation CEO Pay Slice and
Firm Value Around the World Federal Reserve Board Working Paper
Cuntildeat V M Gine and M Guadalupe 2012 The Vote is Cast The Effect of Corporate
Governance on Shareholder Value Journal of Finance 671943-1977
Cuntildeat V M Gine and M Guadalupe 2013 Say Pays Shareholder Voice and Firm
Performance ECGI - Finance Working Paper No 373
Del Guercio D L Seery and T Woidtke 2008 Do Boards Pay Attention When Institutional
Investors lsquoJust Vote Norsquo Journal of Financial Economics 9084-103
Ertimur Y F Ferri and S Stubben 2010 Board of Directorslsquo Responsiveness to Shareholders
Evidence from Shareholder Proposals Journal of Corporate Finance 1653-72
Ertimur Y F Ferri and V Muslu 2011 Shareholder Activism and CEO Pay Review of
Financial Studies 24535ndash592
Ertimur Y F Ferri and D Oesch 2013 Shareholder Votes and Proxy Advisors Evidence from
Say on Pay Journal of Accounting Research 51951-996
Ertimur Y F Ferri and D Oesch forthcoming Does the Director Election System Matter
Evidence from Majority Voting Review of Accounting Studies forthcoming
Ferri F 2012 Low-Cost Shareholder Activism A Review of the Evidence Research
Handbook on the Economics of Corporate Law ed CA Hill and BH McDonnell
Edward Elgar Publishing
Ferri F and D Maber 2013 Say on Pay Votes and CEO Compensation Evidence from the UK
Review of Finance 17527-563
Ferri F and D Oesch 2013 Management Influence on Investors Evidence from Shareholder
Votes on the Frequency of Say on Pay Columbia Business School Research Paper No 13-17
Ferri F and T Sandino 2009 The impact of shareholder activism on financial reporting and
compensation The case of employee stock options expensing The Accounting Review 84433ndash
466
Ferri F and J Weber 2009 AFSCME vs Mozilohellipand Say on Pay for All (A) Harvard
Business School Case 109-009
Gompers P J Ishii and A Metrick 2003 Corporate Governance and Equity Prices Quarterly
Journal of Economics 118107-155
Gordon J 2009 Say on Paylsquo Cautionary Notes on the UK Experience and the Case for
Shareholders Opt-In Harvard Journal on Legislation 46323-368
Iliev P and S Vitanova 2013 The effect of the Say on Pay in the US Pennsylvania State
University Working Paper
Kaplan S 2007 Testimony of Steven N Kaplan on Empowering Shareholders on Executive
Compensation and HR 1257 the Shareholder Vote on Executive Compensation Act before the
Committee on Financial Services United States House of Representatives March 8
Larcker D A McCall and G Ormazabal 2013 The Economic Consequences of Proxy
Advisor Say-on-Pay Voting Policies Working Paper Stanford University
Larcker D G Ormazabal and D Taylor 2011 The market reaction to corporate governance
regulation Journal of Financial Economics 101431ndash448
Levit D and N Malenko 2011 Non-binding voting for shareholder proposals Journal of
Finance 661579ndash1614
Mishel L and N Sabadish 2012 How executive compensation and financial-sector pay have
fueled income inequality Issue Brief 331 Economic Policy Institute
Murphy K 2012 The Politics of Pay A Legislative History of Executive Compensation in
Jennifer Hill and Randall Thomas eds Research Handbook on Executive Pay Edward Elgar
Publishers
New York Times 2007 House votes to give investors say on executive pay April 21
Norris F 2004 Corporate democracy and the power to embarrass New York Times March 4
Ozbas O and J Matsusaka2013 Managerial Accommodation Proxy Access and the Cost of
Shareholder Empowerment University of Southern California CLEO Research Paper Series No
C12-1
Thomas R and J Cotter 2007 Shareholder proposals in the new millennium Shareholder
support board response and market reaction Journal of Corporate Finance 13 368ndash391
Thomas R A Palmiter and J Cotter 2012 Dodd-Franks Say on Pay Will it Lead to a Greater
Role for Shareholders in Corporate Governance Cornell Law Review 971213-1266
Thomas R and C Van der Elst 2013 The International Scope of Say on Pay ECGI Law
Working Paper No227
Introduction
On July 21 2010 President Obama signed into law the Dodd-Frank Wall Street Reform and
Consumer Protection Act (hereafter Dodd-Frank Act)1 One provision of the Act mandates that
beginning with annual meetings on or after January 21 2011 US publicly traded firms allow
shareholders a nonbinding vote on executive pay known as ldquosay on payrdquo (hereafter SOP)2
The purpose of this chapter is to present key insights from the academic research on the
economic consequences of SOP The chapter is organized around five parts Section I provides a
brief history of SOP highlighting its role in the broader context of a gradual shift toward greater
shareholder empowerment The rich set of legislative events accompanying the adoption of SOP
in many countries has provided fertile ground for empirical studies Section II and Section III
review the evidence in such studies focusing on the effect of SOP respectively on executive
pay and firm value followed by some concluding remarks in Section IV3
1 Dodd-Frank Wall Street Reform and Consumer Protection Act Pub L No 111- 203 sect951 124 Stat 1376 1899
(2010) (ldquoDodd-Frank Section 951Prime) Dodd-Frank Section 951 amended the Securities Exchange Act of 1934 by
adding Section 14A (codified as amended at 15 USC sect78n-1) (ldquoExchange Act Section 14Ardquo)
2 Dodd-Frank Section 951 also mandates a non-binding vote on the frequency of future say on pay votes (known as
say-when-on-pay vote) with a choice between an annual a biennial or a triennial frequency This vote on the
frequency of future SOP votes must be held every six years At most firms shareholders overwhelmingly favored an
annual frequency For an analysis of determinants and consequences of say-when-on-pay votes see Ferri and Oesch
(2013)
3 Some of the studies on SOP also examine the determinants of SOP votes Since by and large these studies show
that the key determinants are the same as for other types of shareholder votes (eg proxy advisorsrsquo
recommendations ownership composition performance) I devote little time to this topic and refer the reader to the
Chapter on the Theory and Practice of Corporate Voting by Paul H Edelman and Randall S Thomas
A caveat the academic research on SOP is growing rapidly as more data become available over
time and across countries As a result many studies cited here are in the form of working papers
and their findings should be viewed as preliminary Also I may have missed some of the most
recent studies Finally I apologize if I do not discuss in equal depth all the studies and tend to
focus instead on the work I am more familiar with including my own
I A Brief History of Say on Pay
While CEO pay has made headlines and captured the attention of politicians and policy makers
for many decades (Murphy 2012) it has arguably received even greater scrutiny over the last two
decades During the New Economy of the 1990s the growing use of stock options as incentive
tool (favored by a benign accounting treatment) led to a rapid increase in CEO pay with the
average CEO-to-worker pay ratio peaking at more than 400 in 2000 up from 18 in 1965 (Mishel
and Sabadish 2012) As the dot-com bubble began to burst and a series of accounting and
governance scandals unfolded (eg Worldcom Enron) stock options and executive pay were
blamed for providing perverse incentives to manipulate financial reports and the stock price At
the same time these governance scandals led many institutional investors to take a more active
role in monitoring corporations spurring a wave of shareholder activism A whole new industry
of governance experts and intermediaries (governance ratings agencies proxy advisors)
emerged while influential academic studies documented a large impact of governance quality on
firm value (Gompers Ishii and Metrick 2003 Bebchuk Cohen and Ferrell 2009) Calls for
policy reforms empowering shareholders became louder and corporate governance took center
stage in the policy debate (Bebchuk 2005) Like other institutional investors union pension
funds began to take a more active role Because union pension funds tend to be well diversified
holding very small stakes in thousands of firms they could not exert their influence by buying
large stakes in firms Hence they resorted to lsquolow-costrsquo tools of activism (Ferri 2012) such as
shareholder proposals and shareholder votes on uncontested director elections Tellingly the
percentage of shareholder proposals filed by union pension funds increased from 136 in 1997
to 436 in 2003 when union pension funds surpassed individual investors in terms of number
of proposals and continued to grow thereafter (Ertimur Ferri and Stubben 2010) Not
surprisingly executive pay became a central focus of unionsrsquo governance activism Between
2003 and 2010 led by union pension fundsrsquo efforts the frequency of and voting support for
compensation-related shareholder proposals and compensation-related lsquovote-norsquo campaigns
against directors up for election increased quickly For example Ertimur Ferri and Muslu (2011)
report that among SampP 1500 firms there were approximately 66 proposals per year in the 1997-
2002 period (averaging 162 votes in favor) compared to about 160 proposals per year in the
2003-2007 period (289 votes in favor) New types of shareholder proposals emerged (eg
proposals to introduce performance-based vesting conditions in equity grants expense stock
options subject severance payments to shareholder approval) gaining higher voting support
While these votes were nonbinding firms began to respond particularly when votes were
withheld from compensation committee members up for re-election (Del Guercio Seery and
Woidtke 2008 Cai Garner and Walkling 2009 Ertimur Ferri and Muslu 2011)
Meanwhile an important development had taken place in the United Kingdom (UK) where in
2002 the UK government had introduced a mandatory annual advisory vote on executive
compensation known as ldquosay on payrdquo largely in response to a perceived increase in US-style
lsquofat catrsquo executive pay packages among UK firms During the first say on pay proxy season a
failed say on pay vote at Glaxo SmithKline (with 507 of the votes cast against approval of the
remuneration report) made headlines around the world Shareholders had objected to an
estimated pound22 million severance arrangement for the CEO (reflecting a two-year notice period)
lack of challenging performance targets and the presence of a retesting provision in the stock
option plan4 Glaxo SmithKlinersquos board responded to the vote by launching an extensive
consultation process with shareholders and adjusting the compensation package accordingly
(Ferri and Maber 2013)
In the United States the American Federation of State County and Municipal Employees a
union pension fund took notice of the UK experience and rallied other activist investors to
submit shareholder proposals requesting the adoption of SOP (Ferri and Weber 2009) Between
2006 and 2010 SOP proposals were submitted and voted upon at hundreds of US firms
averaging more than 43 votes in favor and often winning a majority vote with increasing
success over the years (Cuntildeat Gine and Guadalupe 2013) High-profile executive pay scandals
(eg option backdating the large severance package awarded to Nardelli at Home Depot) helped
activists in making their case for more ldquosay on payrdquo among investors and policy makers While
submitted at individual companies the objective of the SOP proposals was to induce widespread
adoption of SOP either via voluntary adoption (which turned out to be rare) a listing standard
or regulatory intervention Policy makers took notice and on April 20 2007 the House of
Representatives passed a bill requiring a nonbinding annual shareholder vote on executive
4 Retesting provisions in the performance-based vesting conditions of equity grants allow to reevaluate in
subsequent years performance targets not achieved during the initial measurement period rather than allowing the
equity grant to lapse As such they were criticized in the UK as a form of lsquoreward for failurersquo
compensation (hereafter SOP Bill)5 On the same day then-Senator Barack Obama introduced a
companion bill in the Senate (S1181)
But it was the financial crisis of 2007-2008 to accelerate the path toward mandatory
adoption of SOP Growing income inequality public outrage over Wall Street excesses (eg the
large bonuses paid at AIG) and banksrsquo bailouts and the perception that executive pay played a
role in inducing excessive risk-taking pressured policy-makers to take action with respect to
executive pay Many of the legislative proposals discussed in the House and Senate in 2008 and
2009 contained a SOP provision (see Table 1 in Larcker Ormazabal and Taylor 2011) Also
holding a SOP vote was a mandatory condition for firms to receive funds under the Troubled
Asset Relief Program (TARP) During the 2008 campaign both Presidential candidates
expressed support for SOP As these events unfolded the debate on merits and drawbacks of
SOP heated up Critics of SOP argued that at best SOP votes would be ignored (because
nonbinding) and at worst would cause directors to pander to shareholders with special interests
or lacking the required expertise and sophistication ultimately resulting in the adoption of
suboptimal pay practices (Kaplan 2007 Bainbridge 2008)6 They also cautioned that the high
5 House Bill 1257 Shareholder Vote on Executive Compensation Act
6 Analytical studies have also raised questions on the informativeness of SOP votes Ozbas and Matsusaka (2013)
note that SOP votes only reveal shareholder preferences for the current compensation plan compared to the
alternative plan that will be adopted if the current plan is rejected but such alternative plan is unclear in practice
casting doubts on whether the shareholder votes on SOP may be truly informative Levit and Malenko (2011) further
suggest that shareholder votes may not be informative about shareholder preferences when shareholders ignore their
own information and vote strategically conditioning on the chance of casting a pivotal vote
cost of analyzing executive pay at thousands of firms would lead shareholders to outsource
voting decisions to proxy advisors who in turn would minimize their own costs by promoting
one-size-fits-all compensation practices which would hurt firm value Finally SOP opponents
noted that shareholders had already tools to express their views on compensation matters such as
the ability to submit shareholder proposals on compensation and withhold votes from directors
responsible for compensation packages Supporters of SOP argued that enhanced shareholder
voice (as formalized in a SOP vote) and reputation concerns would help boards overcome
psychological barriers to negotiating with CEOs on behalf of shareholders resulting in more
efficient compensation contracts and more dialogue between boards and shareholders (Bebchuk
2007) They also pointed out to growing evidence of shareholdersrsquo sophistication in casting
informed votes Finally they argued that SOP would be more effective than shareholder
proposals (limited to a single issue) and less disruptive than a confrontational vote against an
otherwise valuable director More fundamentally critics and supporters of SOP disagreed on
whether existing compensation contracts were the result of an efficient labor market for talent or
instead the expression of management power over captive boards
Finally in 2010 a SOP provision found its way in the Dodd-Frank Act Ironically SOP a
provision designed to provide more alignment between shareholder and management was
adopted as part of a financial reform package aimed at deterring lsquoexcessiversquo risk taking
(excessive from the point of view of social welfare) which may be the result of the alignment of
interests between shareholders and management
As of today non-binding SOP votes (on the compensation report) have been mandated in
Australia Belgium Denmark Israel Italy Portugal while Netherlands Sweden Norway South
Africa and Switzerland adopted some version of a binding SOP vote (on the future compensation
policy) which is now also being considered by the European Union7
When putting SOP into its historical context it is also important to appreciate that SOP has come
to epitomize a broader movement toward greater shareholder democracy beyond the issue of
executive pay For example during the same period shareholder activists have long
(unsuccessfully) lobbied for a proxy access rule that would have made it easier for shareholders
to oust corporate directors and nominate their own candidates Hence many observers often
view the successes and failures of SOP as speaking to the potential effects of other reforms
aimed at increasing shareholder power making the academic research on SOP all more relevant
II The Effect of Say on Pay on Executive Compensation
A Effect of SOP votes on level and composition of executive pay
As discussed earlier United Kingdom was the first country to adopt SOP in 2002 for publicly
traded firms Ferri and Maber (2013) report that in most cases shareholders voted in favor of
compensation plans Failed SOP votes (ie greater than 50 of votes against) were rare (2 of
the sample) though highly publicized in the press However about one-fourth of the sample
firms received more than 20 of votes against
To capture the overall effect of SOP on CEO pay Ferri and Maber (2013) examine the
sensitivity of CEO pay to its economic determinants over the 2000-2005 period (ie before and
after the introduction of SOP regulation) and document a significant increase in the sensitivity of
7 Some countries have already modified their first SOP legislation In 2013 the UK added to the non-binding
backward-looking SOP vote on the current compensation report a binding forward-looking vote on the proposed
compensation policy In 2012 Australia modified the non-binding SOP vote to introduce the so-called ldquotwo-strikesrdquo
rule Under the rule if 25 of shareholders vote against a companyrsquos remuneration report at two consecutive annual
general meetings the entire board may have to stand for re-election within three months For more details about the
international development of SOP see Thomas and Van der Elst (2013)
CEO pay to poor performance They also find that this increase does not occur for a subset of
UK firms exempted from the SOP regulation (firms traded on the Alternative Investment Market
a sub-market of the London Stock Exchange with a more flexible regulatory system) suggesting
that the result reflects the impact of SOP rather than a general trend affecting all firms Besides
Ferri and Maber (2013) find that the increase is more pronounced in firms experiencing high
voting dissent and firms with high abnormal CEO pay before the adoption of SOP consistent
with a causal impact of SOP In contrast they fail to find any effect of SOP on the level of pay
Correa and Lel (2013) examine the effect of SOP laws on CEO pay using a large cross-country
sample of about 103000 firm-year observations from 39 countries including 12 countries that
adopted some (advisory or binding) version of SOP They essentially compare post-SOP CEO
pay at firms of countries adopting SOP to a control sample which includes all non-SOP
observations (that is pre-SOP CEO pay in countries eventually adopting SOP as well as CEO
pay in countries never adopting SOP) They find that firms in a post-SOP regime (i) exhibit
lower CEO pay levels (but only in countries adopting an advisory SOP vote) a finding driven by
a relative decline in equity awards and limited to CEOs and not the other top executives (ii)
higher pay-performance sensitivity (the authors do not look at positive and negative performance
separately) The current version of the paper does not examine whether these findings are more
pronounced in (or driven by) firms with excess CEO pay before the adoption of SOP and in
firms experiencing adverse SOP votes
The drawback of the research design in Correa and Lel (2013) is that it is not clear whether the
results reflect time-series changes within countries adopting SOP (pre vs post) or difference
between firms in SOP and non-SOP countries For example further tests in the paper suggest
that the lower CEO pay levels for post-SOP observations are not driven by a pay decrease
relative to the pre-SOP period (in fact country-by-country analyses indicate no change in CEO
pay levels after the adoption of SOP) but by the comparison with non-SOP countries Also while
the authors employ reasonable econometric solutions to deal with this issue some concerns
remains as to whether non-SOP countries can be a good benchmark for what would have
happened in SOP countries had SOP not been adopted (in other words if only countries with
problematic CEO practices adopt SOP it is possible that changes in CEO pay reflect factors
leading to the adoption of SOP rather than the effect of SOP laws per se) Notwithstanding these
concerns these findings are intriguing and future versions of this recent study have the potential
to provide more conclusive evidence on the effect of SOP
While in the US SOP was mandated by the Dodd-Frank Act its implementation was left to the
Securities and Exchange Commission (SEC) The final rule issued in January 2011 exempted
lsquosmallrsquo firms (ie firms with a public float below $75 million) from the SOP provision for two
years Iliev and Vitanova (2013) exploit this setting to estimate the effect of SOP on CEO
compensation In particularly using a regression discontinuity design they compare changes in
CEO pay of firms just above the SEC-imposed threshold to changes in CEO pay of firms just
below the threshold (but otherwise quite similar) They find no differences in terms of level and
composition of CEO pay and in terms of golden parachutes concluding that SOP had no impact
on CEO pay While the identification strategy chosen by the authors has the benefit of reducing
endogeneity issues it comes with the price of focusing only on fairly small firms (those around
the $75 million public float threshold) where CEO pay may not be a problem in the first place
(in which case the lack of impact of SOP would not be surprising but would not speak to its
impact at large firms traditionally the target of compensation-related criticism)
Cuntildeat Gine and Guadalupe (2013) also use a regression discontinuity design but in a different
setting In particular they analyze nonbinding shareholder proposals to adopt SOP (250
proposals between 2006 and 2010) Comparing the changes in CEO pay for firms voluntarily
adopting SOP in response to the proposals to firms not adopting SOP would be difficult since
the adoption decision is endogenous Similarly it would be difficult to compare changes in CEO
pay between firms where the proposal is passed (and thus significantly more likely to be
adopted) and firms where the proposal fails to be approved since the voting outcome is
endogenous as well
Hence Cuntildeat Gine and Guadalupe (2013) use instead a regression discontinuity design that
compares firms where the proposals receive slightly more than 50 of the votes (the threshold
for approval) to otherwise similar firms where the proposals receive slightly less than 50
Using this approach they conclude that the passing and adoption of SOP proposals is not
associated with changes in the level of CEO pay nor its composition and structure (mix of cash
and equity equity incentives level of perks and deferred pay) In some sense this finding
complements the one in Iliev and Vitanova (2013) in that it is based on large firms (shareholder
proposals are typically submitted at large SampP 500 firms) However both Cai and Walkling
(2011) and Cuntildeat Gine and Guadalupe (2013) show that firms targeted by SOP proposals do not
exhibit excess CEO pay or weaker governance8 Hence the lack of an effect at these firms may
8 Activists submitting SOP proposals chose to focus on a broad sample of large firms rather than target only firms
with problematic CEO pay practices to obtain visibility and show to policy-makers that investorrsquos support for SOP
was not limited to problematic firms (Ferri and Weber 2009)
not be surprising More generally the findings cannot be necessarily generalized to firms away
from the 50 threshold or to firms not targeted by SOP shareholder proposals
B Effect of SOP votes on compensation practices
Analyses of the effect of SOP on level and composition of CEO pay may not capture a number
of other changes to compensation contracts that may be induced by SOP votes but are not
reflected in measure of CEO pay For example the introduction of performance-based vesting
provisions is typically not reflected in the estimates of fair value of equity grants used by
researchers in measuring CEO pay Many other changes (eg terms of severance packages) will
only be reflected in measures of CEO pay contingent upon the occurrence of certain events
Hence it is important to complement the evidence in section IIA with an analysis of firmsrsquo
disclosures of changes to observable provisions of compensation contracts Another benefit of
this approach is that it captures the changes that boards explicitly present as a response to SOP
votes and thus it acts as a reality check on regression-based inferences For example a
regression-based finding of say a decrease in CEO pay level would be more credible if
supported by evidence that firms disclose a decision to reduce target levels of CEO pay in
response to an adverse SOP vote (an action that presumably they have an incentive to disclose to
get rewarded by shareholders in subsequent votes)
Two studies collect data on specific changes made to compensation contracts explicitly in
response to SOP votes in the UK and the US using firmsrsquo disclosures in their proxy statements
Ferri and Maber (2013) examine the compensation reports of a sample of UK firms before and
after the first SOP vote and find that firms experiencing higher voting dissent were significantly
more likely to remove compensation provisions criticized by investors as lsquorewards for failurersquo
relative to a matched sample of firms experiencing lower dissent For example they report that
among firms with long notice periods (implying larger severance payments ) the percentage of
high dissent firms that shortened them after the vote (800) was significantly higher than before
the vote (200) and also significantly higher than among low dissent firms after the vote
(333) They report similar findings for another controversial practice namely the presence of
retesting provisions in the performance-based vesting conditions of equity grants Most firms
indicate that these changes were the result of consultations with their major institutional
investors Firmsrsquo responsiveness to SOP votes was lsquorewardedrsquo with a substantial increase in
favorable SOP votes at the subsequent annual meeting Ferri and Maber (2013) also find that
many firms experiencing low voting dissent at the first SOP vote had removed these practices
before the vote9 highlighting the importance of accounting for ex ante effects when assessing the
impact of a new regulation
Ertimur Ferri and Oesch (2013) examine the effect of SOP on compensation practices in the US
in 2011 (first proxy season under mandatory SOP) Their key findings are generally similar to
the evidence from the UK (i) failed SOP votes are rare (about 2 of the sample) though cases
of substantial dissent are more frequent (ii) more than 55 of the firms experiencing significant
voting dissent respond by making material changes to their compensation plan during the
subsequent year (eg introduction of performance-based vesting conditions in equity grants use
of tougher performance targets removal of perks and tax gross-ups removal of controversial
provisions from severance contracts etc) usually in consultation with major institutional
9 In their sample 700 of low dissent firms shortened their notice periods during the year before the SOP vote
investors (iii) firms making changes to their compensation plans receive greater voting support
at the following SOP vote
Ertimur Ferri and Oesch (2013) also highlight the strong influence of the recommendations
released by proxy advisors (particularly Institutional Shareholder Services ISS) with a negative
recommendation being associated with 25 more votes against the compensation plan Further
evidence of the significant influence of ISS is the striking discontinuity in the relation between
the extent of SOP voting dissent and firmsrsquo responsiveness to the vote the percentage of firms
making compensation changes in response to SOP votes jumps from 32 to 72 around the
30 voting dissent threshold Why After the 2011 proxy season ISS had indicated that firms
failing to ldquoadequatelyrdquo respond to SOP voting dissent above 30 would receive a negative
recommendation in 2012 on the SOP proposal and on the election of compensation committee
members
Finally similar to the UK experience there is also anecdotal evidence of analogous
compensation changes being made by US firms ahead of the SOP vote to avoid a negative
proxy advisor recommendation and an adverse shareholder vote (Thomas Palmiter and Cotter
2013 Larcker McCall and Ormazabal 2013)
III The Effect of Say on Pay on Firm Value
In this section we examine the empirical studies on the overall effect of SOP on firm value
A Event studies around the adoption of Say on Pay regulations
As mentioned in the Introduction on April 20 2007 the House of Representatives passed a SOP
Bill by a 2-1 margin On the same day then-Senator Barack Obama introduced a companion
bill (S1181) in the Senate (which was then put on hold by the Senate Banking Committee)
While the SOP Billrsquos approval was expected (Democrats were in control of the House and
supported the SOP Bill) the 2-1 margin was unexpected suggesting some support for SOP
among Republicans
Cai and Walkling (2011) examine the market reaction to the Housersquos approval of the SOP Bill
for a sample of firms in the SampP 1500 index and document a positive reaction for firms more
likely to benefit from greater shareholder voice over executive pay namely firms with high
abnormal CEO cash pay (defined as the difference between actual CEO cash pay and the level
predicted based on known economic determinants such a size performance industry) firms
with low pay-for-performance sensitivity firms with a history of shareholders willing to vote
against compensation-related management proposals and firms with a history of responsiveness
to shareholder pressure on compensation issues In other words they find a positive reaction in
firms where the compensation problems are more severe and a say on pay vote is more likely to
trigger a response (eg more shareholders willing to vote against management and greater firmsrsquo
propensity to respond to an adverse vote) Note however that they do not find a significant
impact for firms with high abnormal equity and total CEO pay
Larcker Ormazabal and Taylor (2011) also examine the market reaction to the Housersquos approval
of the SOP Bill (in addition to other regulatory events related to executive pay and other
governance provisions) Similar to Cai and Wakling (2011) they fail to find a significant impact
for firms with high abnormal total CEO pay (they do not examine the price reaction for firms
with abnormal cash CEO pay or firms with lower pay-performance sensitivity)
A concern with both studies is that the Housersquos approval of a bill does not guarantee passage in
the Senate and approval by the White House In fact the press at the time reported that the
prospects of the bill in the Senate were uncertain (New York Times 2007) and the Bush White
House openly opposed the Bill (Associated Press 2007) Hence it is not clear the extent to
which this event increased the likelihood of a say on pay legislation Arguably as noted in
Section I the only lsquoeventrsquo that substantially increased the likelihood of say on pay legislation
was the financial crisis and the related perception of compensation abuses (eg outrage over AIG
bonuses)
Ferri and Maber (2013) argue that the announcement of the submission of SOP regulation to the
Parliament in the United Kingdom in June 2002 offers a more powerful setting for an event
study since it was largely unexpected and increased substantially the probability of SOP
adoption (Parliamentrsquos approval was virtually guaranteed) Using this event they document
positive abnormal returns for firms with excess CEO pay combined with poor performance and
for firms with controversial pay practices that weaken the penalties for poor performance (those
practices were often removed in response to SOP votes as discussed earlier in Section IIB)
They interpret their findings as consistent with shareholders perceiving SOP as a value
enhancing monitoring mechanism for firms with low pay-to-poor-performance sensitivity
Finally a recent study by Iliev and Vitanova (2013) examines the market reaction around the
announcement of the SEC final SOP rule that on January 25 2011 exempted small firms (ie
firms with a public float below $75 million) from adopting SOP for two years (press reports at
the time noted that the exemption could become permanent) An appealing feature of this setting
is that the SEC decision was an unexpected reversal from the rule proposed in October 2010 rule
under which no publicly traded firm would be exempted
Iliev and Vitanova (2013) compare the market reaction to this announcement for firms just above
the SEC-imposed threshold and thus subject to the SOP rule and firms just below the threshold
(but otherwise quite similar) and thus exempted from SOP While they find not significant
abnormal returns for either group the abnormal returns for exempted firms are significantly
more negative a finding that the authors interpret as evidence of a positive value effect of
mandatory SOP While the setting is an interesting one some concerns remain as exempted
firms are fairly small and the magnitude of the executive pay problem (if any) is unlikely to be
large enough to explain the differential market reaction Also the same study finds no effect on
CEO pay level and composition (see Section IIA) Hence the reason for the differential stock
price reaction remains unclear It would be helpful to know if the event study result is driven by
(or stronger for) exempted firms with excessive pay or problematic pay practices
B Event studies around Shareholder Proposals to Adopt Say on Pay
Starting in 2006 shareholder activists led by union pension funds submitted shareholder
proposal to adopt SOP at hundreds of US firms in an attempt to induce voluntary or mandatory
widespread adoption of SOP
Cai and Walkling (2011) examine the market reaction to proxy filings and annual meetings of
113 firms targeted by nonbinding shareholder proposals to adopt SOP between 2006 and 2008
On average they find insignificant returns around both the submission of the proposal (proxy
filing date) and the vote on the proposal (annual meeting date) But they also find that when the
proposals are filed by union pension funds the abnormal returns (still insignificant) are more
negative than when filed by other activists and that when the proposal is defeated the abnormal
returns (while insignificant) are more positive than when the proposal is passed (it is not clear
whether this result is driven by union-sponsored proposals) They interpret these findings as
evidence that the market views SOP proposals filed by union pension funds are driven by special
interests rather than value maximization10
However caution is required in interpreting this
evidence Virtually all activists filing SOP proposals were coordinated by a group of investors
(mostly but not only union pension funds) who made available to other activists a template for
SOP proposals and a list of target firms (Ferri and Weber 2009) Hence the distinction between
union and non-union proponents in the context of SOP proposals may be only apparent More
importantly the list of target firms was publicly available months before the proxy filing dates
so it is not clear whether the proxy statements contained any new information Similarly the
voting outcome of the SOP proposals may largely be anticipated (based on the composition of
institutional owners proxy advisorsrsquo recommendations etc) Finally the analyses do not control
for other (potentially new) information contained in proxy statements (eg executive
compensation report other items up for a vote at the annual meeting) or other events occurring at
the annual meeting (eg shareholder votes on other items)
Cuntildeat Gine and Guadalupe (2013) also examine the market reaction to the voting outcome of
nonbinding shareholder proposals to adopt SOP but to alleviate these concerns they employ a
lsquofuzzyrsquo regression discontinuity design (RDD) essentially comparing the stock price reaction to
SOP proposals that pass by a small margin to the reaction to SOP proposals that fail by a small
10
Consistent with this interpretation the authors also find that SOP proposals generally targeted larger firms rather
than firms with excessive pay or lower pay-performance sensitivity However as noted by Ferri and Sandino (2009)
activists typically submit shareholder proposals aimed at promoting policy reforms at a broad sample of large firms
rather than the firms that would most benefit from the proposals because they believe that showing widespread
support for the proposal across all types of firms enhances its credibility with policy makers
margin (Cuntildeat Gine and Guadalupe 2012 Ertimur Ferri and Oesch forthcoming) The underlying idea
is that firms around the threshold are likely to have similar characteristics (indeed they do as the
authors show) but differ in the likelihood of implementation which is typically much higher for
proposals passing the threshold (Ertimur Ferri and Stubben 2010 Thomas and Cotter 2007)
Indeed the authors show that the likelihood of subsequently adopting SOP is 48 higher when
the SOP proposal passes by a small margin relatively to when it fails by a small margin Besides
because in these close-call situations the voting outcome is uncertain its resolution (pass or fail)
is likely to convey new information about the likelihood of SOP adoption making this setting
suitable to an event study11
Using this RDD approach Cuntildeat Gine and Guadalupe (2013) find that on the day of the vote a
SOP proposal that passes by a small margin yields an abnormal return of 24 relative to one
that fails (after controlling for other proposals voted upon at the same meeting) and estimate the
lsquofullrsquo value of SOP at 46 (after taking into account the increase in the probability of SOP
implementation)12
Aside from the issues discussed in Section IA (ie generalizability to other
firms) this estimate seems too large to reflect the present value of future reductions in excess
CEO pay (as noted by the authors their estimate of the value of SOP is equivalent to the
estimated value of removing two anti-takeover provision based on Cuntildeat Gine and Guadalupe
2012) particularly because (i) the authors find no evidence of subsequent changes in levels and
composition of CEO pay for firms implementing SOP (see Section IIA) an (ii) the sample firms
11
Consistent with this observation the authors find no market reaction to the voting outcome of SOP shareholder
proposals away from the threshold
12 Since the outcome of the vote is not binding the 24 market reaction only reflects the expected increase in the
probability of SOP adoption and thus understates the value of the SOP provision
do not appear to be characterized by excess CEO pay However the authors also find significant
improvements in subsequent operating performance and efficiency as a result of passing SOP
proposals and thus conjecture that the SOP vote is viewed as a tool to express a vote of
confidence in management performance more than a way to change compensation practices and
that the large positive market reaction reflects expected performance improvements under this
tighter monitoring regime While this is an interesting idea it remains unclear what additional
ldquoteethrdquo a SOP vote may provide over existing (non-compensation related) monitoring
mechanisms shareholders can express (and do often express) their dissatisfaction with
management performance when voting on director elections (Cai Garner and Walkling 2009)
C Event studies around compensation changes induced by SOP
Two studies examine announcements of changes to compensation plans related to SOP Larcker
McCall and Ormazabal (2013) examine the market reaction to (non-contaminated) compensation
changes disclosed in 8-K filings by Russell 3000 firms in the US during the year before the first
SOP vote They find that changes that appear to be made to avoid a negative proxy advisorrsquos
recommendation and thus a negative SOP vote are associated with a negative abnormal return
of -044 whereas other compensation changes are not associated with a significant stock price
reaction Ertimur Ferri and Oesch (2013) perform a similar analysis but focusing on
compensation changes made after the first SOP vote (and explicitly in response to the vote) by
firms receiving a negative recommendation and a high voting dissent in 2011 They fail to find
significant abnormal returns even for the subset of compensation changes that resulted in a
positive recommendation and low dissent in 2012 (and thus were presumably perceived to be
adequate and material by proxy advisors and voting shareholders)
D Other approaches effect of SOP on Tobinrsquos Q
In their cross-country study Correa and Lel (2013) also examine the effect of SOP laws on
Tobinrsquos Q and report a 36 increase in firm value following the adoption of the SOP laws
(more precisely a 36 higher firm value for post-SOP observations relative to a control sample
that pools together pre-SOP and non-SOP observations) They acknowledge that this increase in
firm value is too large to be justified by the relative decrease in CEO pay that they document and
suggest (but do not test) that it may reflect better alignment of pay and performance13
A key
concern is that higher valuation in countries with SOP laws may reflect other governance
changes introduced at the same time While the study controls for other compensation-related
laws there may be other non-compensation regulations introduced with SOP and with a
potentially larger impact on firm value
IV What have we learned
Adoption of SOP in the US has been long advocated by those who contend that executive pay is
a manifestation of rather than a solution to the agency problem and by those who favor great
shareholder involvement in corporate decisions A growing body of research is starting to
investigate the effect of SOP on executive pay and firm value It is premature to draw definitive
conclusions also because these studies differ in methodologies and settings but three points
seem to emerge First there is robust evidence that boards respond to SOP votes when
13
They also find that the firm value increase is higher when the relative decrease in CEO pay results in a lower pay
differential between CEO and other top managers (CEO pay slice) implying that a source of value creation may be
the reduced pay inequality among the top management team But this seems unlikely to explain the change in
Tobinrsquos Q since the increase occurs for both firms with advisory SOP laws and firms with binding SOP laws and
there is no change in CEO pay slice in firms subject to binding SOP laws
shareholders use it ldquovoicerdquo is heardrdquo Both in the US and UK studies show that not only firms
failing to win the SOP vote but also firms facing substantial dissent (20-30 of the votes
against) make changes to their compensation plans in consultation with institutional investors
and proxy advisors Second in both the US and UK institutional investors appear to use the
power of SOP votes to pressure firms into strengthening the link (or the perceived link) between
pay and performance (particularly on the downside) but not to pressure firms to reduce target
levels of pay suggesting a reluctance of institutional investors to lsquoregulatersquo executive pay Third
and consistent with the above most studies examining the aggregate effect of SOP on CEO pay
find some increase in pay-performance sensitivity but not effect on the level and growth of CEO
pay While some observers may interpret the lack of an effect on CEO pay levels as evidence of
the failure of SOP one should note that SOP per se is a neutral tool Its impact depends on what
investors choose to lsquosay on payrsquo Fourth based on my interpretation of the existing research to
date there is little evidence of an effect of SOP on firm value The studies documenting a
positive price impact appear subject to alternative interpretations or not plausible in their
findings (eg they fail to find an effect on CEO pay or the effect is too small to justify the
documented price impact hence it remains unclear what the source of value creation is) Also
there seems to be no stock price reaction to the SOP-induced compensation changes Finally
given the nature of these SOP-induced changes while some of them may be beneficial it seems
hard to believe that they will have a statistically detectable impact on firm value except perhaps
in a handful of firms where the compensation plan is subject to a complete overhaul
Perhaps one explanation for the weak impact of SOP is that executive pay problems were
overstated or that by the time SOP was introduced (more than a decade after the Enron-type
scandals that led to calls for greater shareholder voice) they had been already addressed via
other mechanisms (hedge fund activism monitoring by institutional investors vote-no
campaigns against compensation committee members SEC-mandated pay disclosures)
Overall though based on the evidence to date it does not appear that SOP had a major impact14
Was the adoption of SOP an lsquooptimalrsquo choice from a social welfare point of view This is a
difficult question to answer given the challenges of identifying and measuring all the potential
costs and benefits associated with SOP (likely to change over time and differ across countries)
Perhaps the most positive view of SOP is that its introduction prevented the adoption of other
more radical and intrusive regulatory measures (eg CEO pay caps) In that sense it may be
viewed as an lsquooptimalrsquo answer (ie the lesser of two evils) to the political pressure to reform
executive pay during the financial crisis
14
This interpretation of the evidence is also consistent with the prediction of analytical studies Ozbas and
Matsusaka (2013) model the benefits and costs of shareholder empowerment distinguishing between the right to
approve and the right to propose They show that permitting shareholders to propose directors or policies can cause
value-maximizing managers to take value-reducing actions to accommodate activist investors with non-value-
maximizing goals As for the right to approve it is weakly beneficial that is approval rights (such as a SOP vote)
are generally beneficial for shareholders in that they limit the managerrsquos ability to pursue private benefits at
shareholder expense but they have minimal impact on managerial actions and firm value (because the manager in
effect can threaten shareholders with an undesirable status quo if they do not approve the managerrsquos proposed
action)
References
Associated Press 2007 Administration opposes say on pay bill April 19
Bainbridge S 2008 Remarks on say on pay An unjustified incursion on director authority
UCLA School of Law Research Paper No 08-06
Bebchuk L 2005 The Case for Increasing Shareholder Power Harvard Law Review 118833-
917
Bebchuk L 2007 Written testimony submitted before the Committee on Financial Services
United States House of Representatives Hearing on Empowering Shareholders on Executive
Compensation March 8
Bebchuk LA A Cohen and A Ferrell 2009 What Matters in Corporate Governance Review
of Financial Studies 22783-827
Cai J J Garner and R Walkling 2009 Electing Directors Journal of Finance 642387-2419
Cai J and R Walkling 2011 Shareholdersrsquo Say on Pay Does it Create Value Journal of
Financial and Quantitative Analysis 46299ndash339
Coates IV JC (2009) Testimony before the Subcommittee on Securities Insurance and
Investment of the Committee on Banking Housing and Urban Affairs United States Senate
July 29 2009 httppapersssrncomsol3paperscfmabstract_id=1475355
Correa R and U Lel 2013 Say On Pay Laws Executive Compensation CEO Pay Slice and
Firm Value Around the World Federal Reserve Board Working Paper
Cuntildeat V M Gine and M Guadalupe 2012 The Vote is Cast The Effect of Corporate
Governance on Shareholder Value Journal of Finance 671943-1977
Cuntildeat V M Gine and M Guadalupe 2013 Say Pays Shareholder Voice and Firm
Performance ECGI - Finance Working Paper No 373
Del Guercio D L Seery and T Woidtke 2008 Do Boards Pay Attention When Institutional
Investors lsquoJust Vote Norsquo Journal of Financial Economics 9084-103
Ertimur Y F Ferri and S Stubben 2010 Board of Directorslsquo Responsiveness to Shareholders
Evidence from Shareholder Proposals Journal of Corporate Finance 1653-72
Ertimur Y F Ferri and V Muslu 2011 Shareholder Activism and CEO Pay Review of
Financial Studies 24535ndash592
Ertimur Y F Ferri and D Oesch 2013 Shareholder Votes and Proxy Advisors Evidence from
Say on Pay Journal of Accounting Research 51951-996
Ertimur Y F Ferri and D Oesch forthcoming Does the Director Election System Matter
Evidence from Majority Voting Review of Accounting Studies forthcoming
Ferri F 2012 Low-Cost Shareholder Activism A Review of the Evidence Research
Handbook on the Economics of Corporate Law ed CA Hill and BH McDonnell
Edward Elgar Publishing
Ferri F and D Maber 2013 Say on Pay Votes and CEO Compensation Evidence from the UK
Review of Finance 17527-563
Ferri F and D Oesch 2013 Management Influence on Investors Evidence from Shareholder
Votes on the Frequency of Say on Pay Columbia Business School Research Paper No 13-17
Ferri F and T Sandino 2009 The impact of shareholder activism on financial reporting and
compensation The case of employee stock options expensing The Accounting Review 84433ndash
466
Ferri F and J Weber 2009 AFSCME vs Mozilohellipand Say on Pay for All (A) Harvard
Business School Case 109-009
Gompers P J Ishii and A Metrick 2003 Corporate Governance and Equity Prices Quarterly
Journal of Economics 118107-155
Gordon J 2009 Say on Paylsquo Cautionary Notes on the UK Experience and the Case for
Shareholders Opt-In Harvard Journal on Legislation 46323-368
Iliev P and S Vitanova 2013 The effect of the Say on Pay in the US Pennsylvania State
University Working Paper
Kaplan S 2007 Testimony of Steven N Kaplan on Empowering Shareholders on Executive
Compensation and HR 1257 the Shareholder Vote on Executive Compensation Act before the
Committee on Financial Services United States House of Representatives March 8
Larcker D A McCall and G Ormazabal 2013 The Economic Consequences of Proxy
Advisor Say-on-Pay Voting Policies Working Paper Stanford University
Larcker D G Ormazabal and D Taylor 2011 The market reaction to corporate governance
regulation Journal of Financial Economics 101431ndash448
Levit D and N Malenko 2011 Non-binding voting for shareholder proposals Journal of
Finance 661579ndash1614
Mishel L and N Sabadish 2012 How executive compensation and financial-sector pay have
fueled income inequality Issue Brief 331 Economic Policy Institute
Murphy K 2012 The Politics of Pay A Legislative History of Executive Compensation in
Jennifer Hill and Randall Thomas eds Research Handbook on Executive Pay Edward Elgar
Publishers
New York Times 2007 House votes to give investors say on executive pay April 21
Norris F 2004 Corporate democracy and the power to embarrass New York Times March 4
Ozbas O and J Matsusaka2013 Managerial Accommodation Proxy Access and the Cost of
Shareholder Empowerment University of Southern California CLEO Research Paper Series No
C12-1
Thomas R and J Cotter 2007 Shareholder proposals in the new millennium Shareholder
support board response and market reaction Journal of Corporate Finance 13 368ndash391
Thomas R A Palmiter and J Cotter 2012 Dodd-Franks Say on Pay Will it Lead to a Greater
Role for Shareholders in Corporate Governance Cornell Law Review 971213-1266
Thomas R and C Van der Elst 2013 The International Scope of Say on Pay ECGI Law
Working Paper No227
A caveat the academic research on SOP is growing rapidly as more data become available over
time and across countries As a result many studies cited here are in the form of working papers
and their findings should be viewed as preliminary Also I may have missed some of the most
recent studies Finally I apologize if I do not discuss in equal depth all the studies and tend to
focus instead on the work I am more familiar with including my own
I A Brief History of Say on Pay
While CEO pay has made headlines and captured the attention of politicians and policy makers
for many decades (Murphy 2012) it has arguably received even greater scrutiny over the last two
decades During the New Economy of the 1990s the growing use of stock options as incentive
tool (favored by a benign accounting treatment) led to a rapid increase in CEO pay with the
average CEO-to-worker pay ratio peaking at more than 400 in 2000 up from 18 in 1965 (Mishel
and Sabadish 2012) As the dot-com bubble began to burst and a series of accounting and
governance scandals unfolded (eg Worldcom Enron) stock options and executive pay were
blamed for providing perverse incentives to manipulate financial reports and the stock price At
the same time these governance scandals led many institutional investors to take a more active
role in monitoring corporations spurring a wave of shareholder activism A whole new industry
of governance experts and intermediaries (governance ratings agencies proxy advisors)
emerged while influential academic studies documented a large impact of governance quality on
firm value (Gompers Ishii and Metrick 2003 Bebchuk Cohen and Ferrell 2009) Calls for
policy reforms empowering shareholders became louder and corporate governance took center
stage in the policy debate (Bebchuk 2005) Like other institutional investors union pension
funds began to take a more active role Because union pension funds tend to be well diversified
holding very small stakes in thousands of firms they could not exert their influence by buying
large stakes in firms Hence they resorted to lsquolow-costrsquo tools of activism (Ferri 2012) such as
shareholder proposals and shareholder votes on uncontested director elections Tellingly the
percentage of shareholder proposals filed by union pension funds increased from 136 in 1997
to 436 in 2003 when union pension funds surpassed individual investors in terms of number
of proposals and continued to grow thereafter (Ertimur Ferri and Stubben 2010) Not
surprisingly executive pay became a central focus of unionsrsquo governance activism Between
2003 and 2010 led by union pension fundsrsquo efforts the frequency of and voting support for
compensation-related shareholder proposals and compensation-related lsquovote-norsquo campaigns
against directors up for election increased quickly For example Ertimur Ferri and Muslu (2011)
report that among SampP 1500 firms there were approximately 66 proposals per year in the 1997-
2002 period (averaging 162 votes in favor) compared to about 160 proposals per year in the
2003-2007 period (289 votes in favor) New types of shareholder proposals emerged (eg
proposals to introduce performance-based vesting conditions in equity grants expense stock
options subject severance payments to shareholder approval) gaining higher voting support
While these votes were nonbinding firms began to respond particularly when votes were
withheld from compensation committee members up for re-election (Del Guercio Seery and
Woidtke 2008 Cai Garner and Walkling 2009 Ertimur Ferri and Muslu 2011)
Meanwhile an important development had taken place in the United Kingdom (UK) where in
2002 the UK government had introduced a mandatory annual advisory vote on executive
compensation known as ldquosay on payrdquo largely in response to a perceived increase in US-style
lsquofat catrsquo executive pay packages among UK firms During the first say on pay proxy season a
failed say on pay vote at Glaxo SmithKline (with 507 of the votes cast against approval of the
remuneration report) made headlines around the world Shareholders had objected to an
estimated pound22 million severance arrangement for the CEO (reflecting a two-year notice period)
lack of challenging performance targets and the presence of a retesting provision in the stock
option plan4 Glaxo SmithKlinersquos board responded to the vote by launching an extensive
consultation process with shareholders and adjusting the compensation package accordingly
(Ferri and Maber 2013)
In the United States the American Federation of State County and Municipal Employees a
union pension fund took notice of the UK experience and rallied other activist investors to
submit shareholder proposals requesting the adoption of SOP (Ferri and Weber 2009) Between
2006 and 2010 SOP proposals were submitted and voted upon at hundreds of US firms
averaging more than 43 votes in favor and often winning a majority vote with increasing
success over the years (Cuntildeat Gine and Guadalupe 2013) High-profile executive pay scandals
(eg option backdating the large severance package awarded to Nardelli at Home Depot) helped
activists in making their case for more ldquosay on payrdquo among investors and policy makers While
submitted at individual companies the objective of the SOP proposals was to induce widespread
adoption of SOP either via voluntary adoption (which turned out to be rare) a listing standard
or regulatory intervention Policy makers took notice and on April 20 2007 the House of
Representatives passed a bill requiring a nonbinding annual shareholder vote on executive
4 Retesting provisions in the performance-based vesting conditions of equity grants allow to reevaluate in
subsequent years performance targets not achieved during the initial measurement period rather than allowing the
equity grant to lapse As such they were criticized in the UK as a form of lsquoreward for failurersquo
compensation (hereafter SOP Bill)5 On the same day then-Senator Barack Obama introduced a
companion bill in the Senate (S1181)
But it was the financial crisis of 2007-2008 to accelerate the path toward mandatory
adoption of SOP Growing income inequality public outrage over Wall Street excesses (eg the
large bonuses paid at AIG) and banksrsquo bailouts and the perception that executive pay played a
role in inducing excessive risk-taking pressured policy-makers to take action with respect to
executive pay Many of the legislative proposals discussed in the House and Senate in 2008 and
2009 contained a SOP provision (see Table 1 in Larcker Ormazabal and Taylor 2011) Also
holding a SOP vote was a mandatory condition for firms to receive funds under the Troubled
Asset Relief Program (TARP) During the 2008 campaign both Presidential candidates
expressed support for SOP As these events unfolded the debate on merits and drawbacks of
SOP heated up Critics of SOP argued that at best SOP votes would be ignored (because
nonbinding) and at worst would cause directors to pander to shareholders with special interests
or lacking the required expertise and sophistication ultimately resulting in the adoption of
suboptimal pay practices (Kaplan 2007 Bainbridge 2008)6 They also cautioned that the high
5 House Bill 1257 Shareholder Vote on Executive Compensation Act
6 Analytical studies have also raised questions on the informativeness of SOP votes Ozbas and Matsusaka (2013)
note that SOP votes only reveal shareholder preferences for the current compensation plan compared to the
alternative plan that will be adopted if the current plan is rejected but such alternative plan is unclear in practice
casting doubts on whether the shareholder votes on SOP may be truly informative Levit and Malenko (2011) further
suggest that shareholder votes may not be informative about shareholder preferences when shareholders ignore their
own information and vote strategically conditioning on the chance of casting a pivotal vote
cost of analyzing executive pay at thousands of firms would lead shareholders to outsource
voting decisions to proxy advisors who in turn would minimize their own costs by promoting
one-size-fits-all compensation practices which would hurt firm value Finally SOP opponents
noted that shareholders had already tools to express their views on compensation matters such as
the ability to submit shareholder proposals on compensation and withhold votes from directors
responsible for compensation packages Supporters of SOP argued that enhanced shareholder
voice (as formalized in a SOP vote) and reputation concerns would help boards overcome
psychological barriers to negotiating with CEOs on behalf of shareholders resulting in more
efficient compensation contracts and more dialogue between boards and shareholders (Bebchuk
2007) They also pointed out to growing evidence of shareholdersrsquo sophistication in casting
informed votes Finally they argued that SOP would be more effective than shareholder
proposals (limited to a single issue) and less disruptive than a confrontational vote against an
otherwise valuable director More fundamentally critics and supporters of SOP disagreed on
whether existing compensation contracts were the result of an efficient labor market for talent or
instead the expression of management power over captive boards
Finally in 2010 a SOP provision found its way in the Dodd-Frank Act Ironically SOP a
provision designed to provide more alignment between shareholder and management was
adopted as part of a financial reform package aimed at deterring lsquoexcessiversquo risk taking
(excessive from the point of view of social welfare) which may be the result of the alignment of
interests between shareholders and management
As of today non-binding SOP votes (on the compensation report) have been mandated in
Australia Belgium Denmark Israel Italy Portugal while Netherlands Sweden Norway South
Africa and Switzerland adopted some version of a binding SOP vote (on the future compensation
policy) which is now also being considered by the European Union7
When putting SOP into its historical context it is also important to appreciate that SOP has come
to epitomize a broader movement toward greater shareholder democracy beyond the issue of
executive pay For example during the same period shareholder activists have long
(unsuccessfully) lobbied for a proxy access rule that would have made it easier for shareholders
to oust corporate directors and nominate their own candidates Hence many observers often
view the successes and failures of SOP as speaking to the potential effects of other reforms
aimed at increasing shareholder power making the academic research on SOP all more relevant
II The Effect of Say on Pay on Executive Compensation
A Effect of SOP votes on level and composition of executive pay
As discussed earlier United Kingdom was the first country to adopt SOP in 2002 for publicly
traded firms Ferri and Maber (2013) report that in most cases shareholders voted in favor of
compensation plans Failed SOP votes (ie greater than 50 of votes against) were rare (2 of
the sample) though highly publicized in the press However about one-fourth of the sample
firms received more than 20 of votes against
To capture the overall effect of SOP on CEO pay Ferri and Maber (2013) examine the
sensitivity of CEO pay to its economic determinants over the 2000-2005 period (ie before and
after the introduction of SOP regulation) and document a significant increase in the sensitivity of
7 Some countries have already modified their first SOP legislation In 2013 the UK added to the non-binding
backward-looking SOP vote on the current compensation report a binding forward-looking vote on the proposed
compensation policy In 2012 Australia modified the non-binding SOP vote to introduce the so-called ldquotwo-strikesrdquo
rule Under the rule if 25 of shareholders vote against a companyrsquos remuneration report at two consecutive annual
general meetings the entire board may have to stand for re-election within three months For more details about the
international development of SOP see Thomas and Van der Elst (2013)
CEO pay to poor performance They also find that this increase does not occur for a subset of
UK firms exempted from the SOP regulation (firms traded on the Alternative Investment Market
a sub-market of the London Stock Exchange with a more flexible regulatory system) suggesting
that the result reflects the impact of SOP rather than a general trend affecting all firms Besides
Ferri and Maber (2013) find that the increase is more pronounced in firms experiencing high
voting dissent and firms with high abnormal CEO pay before the adoption of SOP consistent
with a causal impact of SOP In contrast they fail to find any effect of SOP on the level of pay
Correa and Lel (2013) examine the effect of SOP laws on CEO pay using a large cross-country
sample of about 103000 firm-year observations from 39 countries including 12 countries that
adopted some (advisory or binding) version of SOP They essentially compare post-SOP CEO
pay at firms of countries adopting SOP to a control sample which includes all non-SOP
observations (that is pre-SOP CEO pay in countries eventually adopting SOP as well as CEO
pay in countries never adopting SOP) They find that firms in a post-SOP regime (i) exhibit
lower CEO pay levels (but only in countries adopting an advisory SOP vote) a finding driven by
a relative decline in equity awards and limited to CEOs and not the other top executives (ii)
higher pay-performance sensitivity (the authors do not look at positive and negative performance
separately) The current version of the paper does not examine whether these findings are more
pronounced in (or driven by) firms with excess CEO pay before the adoption of SOP and in
firms experiencing adverse SOP votes
The drawback of the research design in Correa and Lel (2013) is that it is not clear whether the
results reflect time-series changes within countries adopting SOP (pre vs post) or difference
between firms in SOP and non-SOP countries For example further tests in the paper suggest
that the lower CEO pay levels for post-SOP observations are not driven by a pay decrease
relative to the pre-SOP period (in fact country-by-country analyses indicate no change in CEO
pay levels after the adoption of SOP) but by the comparison with non-SOP countries Also while
the authors employ reasonable econometric solutions to deal with this issue some concerns
remains as to whether non-SOP countries can be a good benchmark for what would have
happened in SOP countries had SOP not been adopted (in other words if only countries with
problematic CEO practices adopt SOP it is possible that changes in CEO pay reflect factors
leading to the adoption of SOP rather than the effect of SOP laws per se) Notwithstanding these
concerns these findings are intriguing and future versions of this recent study have the potential
to provide more conclusive evidence on the effect of SOP
While in the US SOP was mandated by the Dodd-Frank Act its implementation was left to the
Securities and Exchange Commission (SEC) The final rule issued in January 2011 exempted
lsquosmallrsquo firms (ie firms with a public float below $75 million) from the SOP provision for two
years Iliev and Vitanova (2013) exploit this setting to estimate the effect of SOP on CEO
compensation In particularly using a regression discontinuity design they compare changes in
CEO pay of firms just above the SEC-imposed threshold to changes in CEO pay of firms just
below the threshold (but otherwise quite similar) They find no differences in terms of level and
composition of CEO pay and in terms of golden parachutes concluding that SOP had no impact
on CEO pay While the identification strategy chosen by the authors has the benefit of reducing
endogeneity issues it comes with the price of focusing only on fairly small firms (those around
the $75 million public float threshold) where CEO pay may not be a problem in the first place
(in which case the lack of impact of SOP would not be surprising but would not speak to its
impact at large firms traditionally the target of compensation-related criticism)
Cuntildeat Gine and Guadalupe (2013) also use a regression discontinuity design but in a different
setting In particular they analyze nonbinding shareholder proposals to adopt SOP (250
proposals between 2006 and 2010) Comparing the changes in CEO pay for firms voluntarily
adopting SOP in response to the proposals to firms not adopting SOP would be difficult since
the adoption decision is endogenous Similarly it would be difficult to compare changes in CEO
pay between firms where the proposal is passed (and thus significantly more likely to be
adopted) and firms where the proposal fails to be approved since the voting outcome is
endogenous as well
Hence Cuntildeat Gine and Guadalupe (2013) use instead a regression discontinuity design that
compares firms where the proposals receive slightly more than 50 of the votes (the threshold
for approval) to otherwise similar firms where the proposals receive slightly less than 50
Using this approach they conclude that the passing and adoption of SOP proposals is not
associated with changes in the level of CEO pay nor its composition and structure (mix of cash
and equity equity incentives level of perks and deferred pay) In some sense this finding
complements the one in Iliev and Vitanova (2013) in that it is based on large firms (shareholder
proposals are typically submitted at large SampP 500 firms) However both Cai and Walkling
(2011) and Cuntildeat Gine and Guadalupe (2013) show that firms targeted by SOP proposals do not
exhibit excess CEO pay or weaker governance8 Hence the lack of an effect at these firms may
8 Activists submitting SOP proposals chose to focus on a broad sample of large firms rather than target only firms
with problematic CEO pay practices to obtain visibility and show to policy-makers that investorrsquos support for SOP
was not limited to problematic firms (Ferri and Weber 2009)
not be surprising More generally the findings cannot be necessarily generalized to firms away
from the 50 threshold or to firms not targeted by SOP shareholder proposals
B Effect of SOP votes on compensation practices
Analyses of the effect of SOP on level and composition of CEO pay may not capture a number
of other changes to compensation contracts that may be induced by SOP votes but are not
reflected in measure of CEO pay For example the introduction of performance-based vesting
provisions is typically not reflected in the estimates of fair value of equity grants used by
researchers in measuring CEO pay Many other changes (eg terms of severance packages) will
only be reflected in measures of CEO pay contingent upon the occurrence of certain events
Hence it is important to complement the evidence in section IIA with an analysis of firmsrsquo
disclosures of changes to observable provisions of compensation contracts Another benefit of
this approach is that it captures the changes that boards explicitly present as a response to SOP
votes and thus it acts as a reality check on regression-based inferences For example a
regression-based finding of say a decrease in CEO pay level would be more credible if
supported by evidence that firms disclose a decision to reduce target levels of CEO pay in
response to an adverse SOP vote (an action that presumably they have an incentive to disclose to
get rewarded by shareholders in subsequent votes)
Two studies collect data on specific changes made to compensation contracts explicitly in
response to SOP votes in the UK and the US using firmsrsquo disclosures in their proxy statements
Ferri and Maber (2013) examine the compensation reports of a sample of UK firms before and
after the first SOP vote and find that firms experiencing higher voting dissent were significantly
more likely to remove compensation provisions criticized by investors as lsquorewards for failurersquo
relative to a matched sample of firms experiencing lower dissent For example they report that
among firms with long notice periods (implying larger severance payments ) the percentage of
high dissent firms that shortened them after the vote (800) was significantly higher than before
the vote (200) and also significantly higher than among low dissent firms after the vote
(333) They report similar findings for another controversial practice namely the presence of
retesting provisions in the performance-based vesting conditions of equity grants Most firms
indicate that these changes were the result of consultations with their major institutional
investors Firmsrsquo responsiveness to SOP votes was lsquorewardedrsquo with a substantial increase in
favorable SOP votes at the subsequent annual meeting Ferri and Maber (2013) also find that
many firms experiencing low voting dissent at the first SOP vote had removed these practices
before the vote9 highlighting the importance of accounting for ex ante effects when assessing the
impact of a new regulation
Ertimur Ferri and Oesch (2013) examine the effect of SOP on compensation practices in the US
in 2011 (first proxy season under mandatory SOP) Their key findings are generally similar to
the evidence from the UK (i) failed SOP votes are rare (about 2 of the sample) though cases
of substantial dissent are more frequent (ii) more than 55 of the firms experiencing significant
voting dissent respond by making material changes to their compensation plan during the
subsequent year (eg introduction of performance-based vesting conditions in equity grants use
of tougher performance targets removal of perks and tax gross-ups removal of controversial
provisions from severance contracts etc) usually in consultation with major institutional
9 In their sample 700 of low dissent firms shortened their notice periods during the year before the SOP vote
investors (iii) firms making changes to their compensation plans receive greater voting support
at the following SOP vote
Ertimur Ferri and Oesch (2013) also highlight the strong influence of the recommendations
released by proxy advisors (particularly Institutional Shareholder Services ISS) with a negative
recommendation being associated with 25 more votes against the compensation plan Further
evidence of the significant influence of ISS is the striking discontinuity in the relation between
the extent of SOP voting dissent and firmsrsquo responsiveness to the vote the percentage of firms
making compensation changes in response to SOP votes jumps from 32 to 72 around the
30 voting dissent threshold Why After the 2011 proxy season ISS had indicated that firms
failing to ldquoadequatelyrdquo respond to SOP voting dissent above 30 would receive a negative
recommendation in 2012 on the SOP proposal and on the election of compensation committee
members
Finally similar to the UK experience there is also anecdotal evidence of analogous
compensation changes being made by US firms ahead of the SOP vote to avoid a negative
proxy advisor recommendation and an adverse shareholder vote (Thomas Palmiter and Cotter
2013 Larcker McCall and Ormazabal 2013)
III The Effect of Say on Pay on Firm Value
In this section we examine the empirical studies on the overall effect of SOP on firm value
A Event studies around the adoption of Say on Pay regulations
As mentioned in the Introduction on April 20 2007 the House of Representatives passed a SOP
Bill by a 2-1 margin On the same day then-Senator Barack Obama introduced a companion
bill (S1181) in the Senate (which was then put on hold by the Senate Banking Committee)
While the SOP Billrsquos approval was expected (Democrats were in control of the House and
supported the SOP Bill) the 2-1 margin was unexpected suggesting some support for SOP
among Republicans
Cai and Walkling (2011) examine the market reaction to the Housersquos approval of the SOP Bill
for a sample of firms in the SampP 1500 index and document a positive reaction for firms more
likely to benefit from greater shareholder voice over executive pay namely firms with high
abnormal CEO cash pay (defined as the difference between actual CEO cash pay and the level
predicted based on known economic determinants such a size performance industry) firms
with low pay-for-performance sensitivity firms with a history of shareholders willing to vote
against compensation-related management proposals and firms with a history of responsiveness
to shareholder pressure on compensation issues In other words they find a positive reaction in
firms where the compensation problems are more severe and a say on pay vote is more likely to
trigger a response (eg more shareholders willing to vote against management and greater firmsrsquo
propensity to respond to an adverse vote) Note however that they do not find a significant
impact for firms with high abnormal equity and total CEO pay
Larcker Ormazabal and Taylor (2011) also examine the market reaction to the Housersquos approval
of the SOP Bill (in addition to other regulatory events related to executive pay and other
governance provisions) Similar to Cai and Wakling (2011) they fail to find a significant impact
for firms with high abnormal total CEO pay (they do not examine the price reaction for firms
with abnormal cash CEO pay or firms with lower pay-performance sensitivity)
A concern with both studies is that the Housersquos approval of a bill does not guarantee passage in
the Senate and approval by the White House In fact the press at the time reported that the
prospects of the bill in the Senate were uncertain (New York Times 2007) and the Bush White
House openly opposed the Bill (Associated Press 2007) Hence it is not clear the extent to
which this event increased the likelihood of a say on pay legislation Arguably as noted in
Section I the only lsquoeventrsquo that substantially increased the likelihood of say on pay legislation
was the financial crisis and the related perception of compensation abuses (eg outrage over AIG
bonuses)
Ferri and Maber (2013) argue that the announcement of the submission of SOP regulation to the
Parliament in the United Kingdom in June 2002 offers a more powerful setting for an event
study since it was largely unexpected and increased substantially the probability of SOP
adoption (Parliamentrsquos approval was virtually guaranteed) Using this event they document
positive abnormal returns for firms with excess CEO pay combined with poor performance and
for firms with controversial pay practices that weaken the penalties for poor performance (those
practices were often removed in response to SOP votes as discussed earlier in Section IIB)
They interpret their findings as consistent with shareholders perceiving SOP as a value
enhancing monitoring mechanism for firms with low pay-to-poor-performance sensitivity
Finally a recent study by Iliev and Vitanova (2013) examines the market reaction around the
announcement of the SEC final SOP rule that on January 25 2011 exempted small firms (ie
firms with a public float below $75 million) from adopting SOP for two years (press reports at
the time noted that the exemption could become permanent) An appealing feature of this setting
is that the SEC decision was an unexpected reversal from the rule proposed in October 2010 rule
under which no publicly traded firm would be exempted
Iliev and Vitanova (2013) compare the market reaction to this announcement for firms just above
the SEC-imposed threshold and thus subject to the SOP rule and firms just below the threshold
(but otherwise quite similar) and thus exempted from SOP While they find not significant
abnormal returns for either group the abnormal returns for exempted firms are significantly
more negative a finding that the authors interpret as evidence of a positive value effect of
mandatory SOP While the setting is an interesting one some concerns remain as exempted
firms are fairly small and the magnitude of the executive pay problem (if any) is unlikely to be
large enough to explain the differential market reaction Also the same study finds no effect on
CEO pay level and composition (see Section IIA) Hence the reason for the differential stock
price reaction remains unclear It would be helpful to know if the event study result is driven by
(or stronger for) exempted firms with excessive pay or problematic pay practices
B Event studies around Shareholder Proposals to Adopt Say on Pay
Starting in 2006 shareholder activists led by union pension funds submitted shareholder
proposal to adopt SOP at hundreds of US firms in an attempt to induce voluntary or mandatory
widespread adoption of SOP
Cai and Walkling (2011) examine the market reaction to proxy filings and annual meetings of
113 firms targeted by nonbinding shareholder proposals to adopt SOP between 2006 and 2008
On average they find insignificant returns around both the submission of the proposal (proxy
filing date) and the vote on the proposal (annual meeting date) But they also find that when the
proposals are filed by union pension funds the abnormal returns (still insignificant) are more
negative than when filed by other activists and that when the proposal is defeated the abnormal
returns (while insignificant) are more positive than when the proposal is passed (it is not clear
whether this result is driven by union-sponsored proposals) They interpret these findings as
evidence that the market views SOP proposals filed by union pension funds are driven by special
interests rather than value maximization10
However caution is required in interpreting this
evidence Virtually all activists filing SOP proposals were coordinated by a group of investors
(mostly but not only union pension funds) who made available to other activists a template for
SOP proposals and a list of target firms (Ferri and Weber 2009) Hence the distinction between
union and non-union proponents in the context of SOP proposals may be only apparent More
importantly the list of target firms was publicly available months before the proxy filing dates
so it is not clear whether the proxy statements contained any new information Similarly the
voting outcome of the SOP proposals may largely be anticipated (based on the composition of
institutional owners proxy advisorsrsquo recommendations etc) Finally the analyses do not control
for other (potentially new) information contained in proxy statements (eg executive
compensation report other items up for a vote at the annual meeting) or other events occurring at
the annual meeting (eg shareholder votes on other items)
Cuntildeat Gine and Guadalupe (2013) also examine the market reaction to the voting outcome of
nonbinding shareholder proposals to adopt SOP but to alleviate these concerns they employ a
lsquofuzzyrsquo regression discontinuity design (RDD) essentially comparing the stock price reaction to
SOP proposals that pass by a small margin to the reaction to SOP proposals that fail by a small
10
Consistent with this interpretation the authors also find that SOP proposals generally targeted larger firms rather
than firms with excessive pay or lower pay-performance sensitivity However as noted by Ferri and Sandino (2009)
activists typically submit shareholder proposals aimed at promoting policy reforms at a broad sample of large firms
rather than the firms that would most benefit from the proposals because they believe that showing widespread
support for the proposal across all types of firms enhances its credibility with policy makers
margin (Cuntildeat Gine and Guadalupe 2012 Ertimur Ferri and Oesch forthcoming) The underlying idea
is that firms around the threshold are likely to have similar characteristics (indeed they do as the
authors show) but differ in the likelihood of implementation which is typically much higher for
proposals passing the threshold (Ertimur Ferri and Stubben 2010 Thomas and Cotter 2007)
Indeed the authors show that the likelihood of subsequently adopting SOP is 48 higher when
the SOP proposal passes by a small margin relatively to when it fails by a small margin Besides
because in these close-call situations the voting outcome is uncertain its resolution (pass or fail)
is likely to convey new information about the likelihood of SOP adoption making this setting
suitable to an event study11
Using this RDD approach Cuntildeat Gine and Guadalupe (2013) find that on the day of the vote a
SOP proposal that passes by a small margin yields an abnormal return of 24 relative to one
that fails (after controlling for other proposals voted upon at the same meeting) and estimate the
lsquofullrsquo value of SOP at 46 (after taking into account the increase in the probability of SOP
implementation)12
Aside from the issues discussed in Section IA (ie generalizability to other
firms) this estimate seems too large to reflect the present value of future reductions in excess
CEO pay (as noted by the authors their estimate of the value of SOP is equivalent to the
estimated value of removing two anti-takeover provision based on Cuntildeat Gine and Guadalupe
2012) particularly because (i) the authors find no evidence of subsequent changes in levels and
composition of CEO pay for firms implementing SOP (see Section IIA) an (ii) the sample firms
11
Consistent with this observation the authors find no market reaction to the voting outcome of SOP shareholder
proposals away from the threshold
12 Since the outcome of the vote is not binding the 24 market reaction only reflects the expected increase in the
probability of SOP adoption and thus understates the value of the SOP provision
do not appear to be characterized by excess CEO pay However the authors also find significant
improvements in subsequent operating performance and efficiency as a result of passing SOP
proposals and thus conjecture that the SOP vote is viewed as a tool to express a vote of
confidence in management performance more than a way to change compensation practices and
that the large positive market reaction reflects expected performance improvements under this
tighter monitoring regime While this is an interesting idea it remains unclear what additional
ldquoteethrdquo a SOP vote may provide over existing (non-compensation related) monitoring
mechanisms shareholders can express (and do often express) their dissatisfaction with
management performance when voting on director elections (Cai Garner and Walkling 2009)
C Event studies around compensation changes induced by SOP
Two studies examine announcements of changes to compensation plans related to SOP Larcker
McCall and Ormazabal (2013) examine the market reaction to (non-contaminated) compensation
changes disclosed in 8-K filings by Russell 3000 firms in the US during the year before the first
SOP vote They find that changes that appear to be made to avoid a negative proxy advisorrsquos
recommendation and thus a negative SOP vote are associated with a negative abnormal return
of -044 whereas other compensation changes are not associated with a significant stock price
reaction Ertimur Ferri and Oesch (2013) perform a similar analysis but focusing on
compensation changes made after the first SOP vote (and explicitly in response to the vote) by
firms receiving a negative recommendation and a high voting dissent in 2011 They fail to find
significant abnormal returns even for the subset of compensation changes that resulted in a
positive recommendation and low dissent in 2012 (and thus were presumably perceived to be
adequate and material by proxy advisors and voting shareholders)
D Other approaches effect of SOP on Tobinrsquos Q
In their cross-country study Correa and Lel (2013) also examine the effect of SOP laws on
Tobinrsquos Q and report a 36 increase in firm value following the adoption of the SOP laws
(more precisely a 36 higher firm value for post-SOP observations relative to a control sample
that pools together pre-SOP and non-SOP observations) They acknowledge that this increase in
firm value is too large to be justified by the relative decrease in CEO pay that they document and
suggest (but do not test) that it may reflect better alignment of pay and performance13
A key
concern is that higher valuation in countries with SOP laws may reflect other governance
changes introduced at the same time While the study controls for other compensation-related
laws there may be other non-compensation regulations introduced with SOP and with a
potentially larger impact on firm value
IV What have we learned
Adoption of SOP in the US has been long advocated by those who contend that executive pay is
a manifestation of rather than a solution to the agency problem and by those who favor great
shareholder involvement in corporate decisions A growing body of research is starting to
investigate the effect of SOP on executive pay and firm value It is premature to draw definitive
conclusions also because these studies differ in methodologies and settings but three points
seem to emerge First there is robust evidence that boards respond to SOP votes when
13
They also find that the firm value increase is higher when the relative decrease in CEO pay results in a lower pay
differential between CEO and other top managers (CEO pay slice) implying that a source of value creation may be
the reduced pay inequality among the top management team But this seems unlikely to explain the change in
Tobinrsquos Q since the increase occurs for both firms with advisory SOP laws and firms with binding SOP laws and
there is no change in CEO pay slice in firms subject to binding SOP laws
shareholders use it ldquovoicerdquo is heardrdquo Both in the US and UK studies show that not only firms
failing to win the SOP vote but also firms facing substantial dissent (20-30 of the votes
against) make changes to their compensation plans in consultation with institutional investors
and proxy advisors Second in both the US and UK institutional investors appear to use the
power of SOP votes to pressure firms into strengthening the link (or the perceived link) between
pay and performance (particularly on the downside) but not to pressure firms to reduce target
levels of pay suggesting a reluctance of institutional investors to lsquoregulatersquo executive pay Third
and consistent with the above most studies examining the aggregate effect of SOP on CEO pay
find some increase in pay-performance sensitivity but not effect on the level and growth of CEO
pay While some observers may interpret the lack of an effect on CEO pay levels as evidence of
the failure of SOP one should note that SOP per se is a neutral tool Its impact depends on what
investors choose to lsquosay on payrsquo Fourth based on my interpretation of the existing research to
date there is little evidence of an effect of SOP on firm value The studies documenting a
positive price impact appear subject to alternative interpretations or not plausible in their
findings (eg they fail to find an effect on CEO pay or the effect is too small to justify the
documented price impact hence it remains unclear what the source of value creation is) Also
there seems to be no stock price reaction to the SOP-induced compensation changes Finally
given the nature of these SOP-induced changes while some of them may be beneficial it seems
hard to believe that they will have a statistically detectable impact on firm value except perhaps
in a handful of firms where the compensation plan is subject to a complete overhaul
Perhaps one explanation for the weak impact of SOP is that executive pay problems were
overstated or that by the time SOP was introduced (more than a decade after the Enron-type
scandals that led to calls for greater shareholder voice) they had been already addressed via
other mechanisms (hedge fund activism monitoring by institutional investors vote-no
campaigns against compensation committee members SEC-mandated pay disclosures)
Overall though based on the evidence to date it does not appear that SOP had a major impact14
Was the adoption of SOP an lsquooptimalrsquo choice from a social welfare point of view This is a
difficult question to answer given the challenges of identifying and measuring all the potential
costs and benefits associated with SOP (likely to change over time and differ across countries)
Perhaps the most positive view of SOP is that its introduction prevented the adoption of other
more radical and intrusive regulatory measures (eg CEO pay caps) In that sense it may be
viewed as an lsquooptimalrsquo answer (ie the lesser of two evils) to the political pressure to reform
executive pay during the financial crisis
14
This interpretation of the evidence is also consistent with the prediction of analytical studies Ozbas and
Matsusaka (2013) model the benefits and costs of shareholder empowerment distinguishing between the right to
approve and the right to propose They show that permitting shareholders to propose directors or policies can cause
value-maximizing managers to take value-reducing actions to accommodate activist investors with non-value-
maximizing goals As for the right to approve it is weakly beneficial that is approval rights (such as a SOP vote)
are generally beneficial for shareholders in that they limit the managerrsquos ability to pursue private benefits at
shareholder expense but they have minimal impact on managerial actions and firm value (because the manager in
effect can threaten shareholders with an undesirable status quo if they do not approve the managerrsquos proposed
action)
References
Associated Press 2007 Administration opposes say on pay bill April 19
Bainbridge S 2008 Remarks on say on pay An unjustified incursion on director authority
UCLA School of Law Research Paper No 08-06
Bebchuk L 2005 The Case for Increasing Shareholder Power Harvard Law Review 118833-
917
Bebchuk L 2007 Written testimony submitted before the Committee on Financial Services
United States House of Representatives Hearing on Empowering Shareholders on Executive
Compensation March 8
Bebchuk LA A Cohen and A Ferrell 2009 What Matters in Corporate Governance Review
of Financial Studies 22783-827
Cai J J Garner and R Walkling 2009 Electing Directors Journal of Finance 642387-2419
Cai J and R Walkling 2011 Shareholdersrsquo Say on Pay Does it Create Value Journal of
Financial and Quantitative Analysis 46299ndash339
Coates IV JC (2009) Testimony before the Subcommittee on Securities Insurance and
Investment of the Committee on Banking Housing and Urban Affairs United States Senate
July 29 2009 httppapersssrncomsol3paperscfmabstract_id=1475355
Correa R and U Lel 2013 Say On Pay Laws Executive Compensation CEO Pay Slice and
Firm Value Around the World Federal Reserve Board Working Paper
Cuntildeat V M Gine and M Guadalupe 2012 The Vote is Cast The Effect of Corporate
Governance on Shareholder Value Journal of Finance 671943-1977
Cuntildeat V M Gine and M Guadalupe 2013 Say Pays Shareholder Voice and Firm
Performance ECGI - Finance Working Paper No 373
Del Guercio D L Seery and T Woidtke 2008 Do Boards Pay Attention When Institutional
Investors lsquoJust Vote Norsquo Journal of Financial Economics 9084-103
Ertimur Y F Ferri and S Stubben 2010 Board of Directorslsquo Responsiveness to Shareholders
Evidence from Shareholder Proposals Journal of Corporate Finance 1653-72
Ertimur Y F Ferri and V Muslu 2011 Shareholder Activism and CEO Pay Review of
Financial Studies 24535ndash592
Ertimur Y F Ferri and D Oesch 2013 Shareholder Votes and Proxy Advisors Evidence from
Say on Pay Journal of Accounting Research 51951-996
Ertimur Y F Ferri and D Oesch forthcoming Does the Director Election System Matter
Evidence from Majority Voting Review of Accounting Studies forthcoming
Ferri F 2012 Low-Cost Shareholder Activism A Review of the Evidence Research
Handbook on the Economics of Corporate Law ed CA Hill and BH McDonnell
Edward Elgar Publishing
Ferri F and D Maber 2013 Say on Pay Votes and CEO Compensation Evidence from the UK
Review of Finance 17527-563
Ferri F and D Oesch 2013 Management Influence on Investors Evidence from Shareholder
Votes on the Frequency of Say on Pay Columbia Business School Research Paper No 13-17
Ferri F and T Sandino 2009 The impact of shareholder activism on financial reporting and
compensation The case of employee stock options expensing The Accounting Review 84433ndash
466
Ferri F and J Weber 2009 AFSCME vs Mozilohellipand Say on Pay for All (A) Harvard
Business School Case 109-009
Gompers P J Ishii and A Metrick 2003 Corporate Governance and Equity Prices Quarterly
Journal of Economics 118107-155
Gordon J 2009 Say on Paylsquo Cautionary Notes on the UK Experience and the Case for
Shareholders Opt-In Harvard Journal on Legislation 46323-368
Iliev P and S Vitanova 2013 The effect of the Say on Pay in the US Pennsylvania State
University Working Paper
Kaplan S 2007 Testimony of Steven N Kaplan on Empowering Shareholders on Executive
Compensation and HR 1257 the Shareholder Vote on Executive Compensation Act before the
Committee on Financial Services United States House of Representatives March 8
Larcker D A McCall and G Ormazabal 2013 The Economic Consequences of Proxy
Advisor Say-on-Pay Voting Policies Working Paper Stanford University
Larcker D G Ormazabal and D Taylor 2011 The market reaction to corporate governance
regulation Journal of Financial Economics 101431ndash448
Levit D and N Malenko 2011 Non-binding voting for shareholder proposals Journal of
Finance 661579ndash1614
Mishel L and N Sabadish 2012 How executive compensation and financial-sector pay have
fueled income inequality Issue Brief 331 Economic Policy Institute
Murphy K 2012 The Politics of Pay A Legislative History of Executive Compensation in
Jennifer Hill and Randall Thomas eds Research Handbook on Executive Pay Edward Elgar
Publishers
New York Times 2007 House votes to give investors say on executive pay April 21
Norris F 2004 Corporate democracy and the power to embarrass New York Times March 4
Ozbas O and J Matsusaka2013 Managerial Accommodation Proxy Access and the Cost of
Shareholder Empowerment University of Southern California CLEO Research Paper Series No
C12-1
Thomas R and J Cotter 2007 Shareholder proposals in the new millennium Shareholder
support board response and market reaction Journal of Corporate Finance 13 368ndash391
Thomas R A Palmiter and J Cotter 2012 Dodd-Franks Say on Pay Will it Lead to a Greater
Role for Shareholders in Corporate Governance Cornell Law Review 971213-1266
Thomas R and C Van der Elst 2013 The International Scope of Say on Pay ECGI Law
Working Paper No227
holding very small stakes in thousands of firms they could not exert their influence by buying
large stakes in firms Hence they resorted to lsquolow-costrsquo tools of activism (Ferri 2012) such as
shareholder proposals and shareholder votes on uncontested director elections Tellingly the
percentage of shareholder proposals filed by union pension funds increased from 136 in 1997
to 436 in 2003 when union pension funds surpassed individual investors in terms of number
of proposals and continued to grow thereafter (Ertimur Ferri and Stubben 2010) Not
surprisingly executive pay became a central focus of unionsrsquo governance activism Between
2003 and 2010 led by union pension fundsrsquo efforts the frequency of and voting support for
compensation-related shareholder proposals and compensation-related lsquovote-norsquo campaigns
against directors up for election increased quickly For example Ertimur Ferri and Muslu (2011)
report that among SampP 1500 firms there were approximately 66 proposals per year in the 1997-
2002 period (averaging 162 votes in favor) compared to about 160 proposals per year in the
2003-2007 period (289 votes in favor) New types of shareholder proposals emerged (eg
proposals to introduce performance-based vesting conditions in equity grants expense stock
options subject severance payments to shareholder approval) gaining higher voting support
While these votes were nonbinding firms began to respond particularly when votes were
withheld from compensation committee members up for re-election (Del Guercio Seery and
Woidtke 2008 Cai Garner and Walkling 2009 Ertimur Ferri and Muslu 2011)
Meanwhile an important development had taken place in the United Kingdom (UK) where in
2002 the UK government had introduced a mandatory annual advisory vote on executive
compensation known as ldquosay on payrdquo largely in response to a perceived increase in US-style
lsquofat catrsquo executive pay packages among UK firms During the first say on pay proxy season a
failed say on pay vote at Glaxo SmithKline (with 507 of the votes cast against approval of the
remuneration report) made headlines around the world Shareholders had objected to an
estimated pound22 million severance arrangement for the CEO (reflecting a two-year notice period)
lack of challenging performance targets and the presence of a retesting provision in the stock
option plan4 Glaxo SmithKlinersquos board responded to the vote by launching an extensive
consultation process with shareholders and adjusting the compensation package accordingly
(Ferri and Maber 2013)
In the United States the American Federation of State County and Municipal Employees a
union pension fund took notice of the UK experience and rallied other activist investors to
submit shareholder proposals requesting the adoption of SOP (Ferri and Weber 2009) Between
2006 and 2010 SOP proposals were submitted and voted upon at hundreds of US firms
averaging more than 43 votes in favor and often winning a majority vote with increasing
success over the years (Cuntildeat Gine and Guadalupe 2013) High-profile executive pay scandals
(eg option backdating the large severance package awarded to Nardelli at Home Depot) helped
activists in making their case for more ldquosay on payrdquo among investors and policy makers While
submitted at individual companies the objective of the SOP proposals was to induce widespread
adoption of SOP either via voluntary adoption (which turned out to be rare) a listing standard
or regulatory intervention Policy makers took notice and on April 20 2007 the House of
Representatives passed a bill requiring a nonbinding annual shareholder vote on executive
4 Retesting provisions in the performance-based vesting conditions of equity grants allow to reevaluate in
subsequent years performance targets not achieved during the initial measurement period rather than allowing the
equity grant to lapse As such they were criticized in the UK as a form of lsquoreward for failurersquo
compensation (hereafter SOP Bill)5 On the same day then-Senator Barack Obama introduced a
companion bill in the Senate (S1181)
But it was the financial crisis of 2007-2008 to accelerate the path toward mandatory
adoption of SOP Growing income inequality public outrage over Wall Street excesses (eg the
large bonuses paid at AIG) and banksrsquo bailouts and the perception that executive pay played a
role in inducing excessive risk-taking pressured policy-makers to take action with respect to
executive pay Many of the legislative proposals discussed in the House and Senate in 2008 and
2009 contained a SOP provision (see Table 1 in Larcker Ormazabal and Taylor 2011) Also
holding a SOP vote was a mandatory condition for firms to receive funds under the Troubled
Asset Relief Program (TARP) During the 2008 campaign both Presidential candidates
expressed support for SOP As these events unfolded the debate on merits and drawbacks of
SOP heated up Critics of SOP argued that at best SOP votes would be ignored (because
nonbinding) and at worst would cause directors to pander to shareholders with special interests
or lacking the required expertise and sophistication ultimately resulting in the adoption of
suboptimal pay practices (Kaplan 2007 Bainbridge 2008)6 They also cautioned that the high
5 House Bill 1257 Shareholder Vote on Executive Compensation Act
6 Analytical studies have also raised questions on the informativeness of SOP votes Ozbas and Matsusaka (2013)
note that SOP votes only reveal shareholder preferences for the current compensation plan compared to the
alternative plan that will be adopted if the current plan is rejected but such alternative plan is unclear in practice
casting doubts on whether the shareholder votes on SOP may be truly informative Levit and Malenko (2011) further
suggest that shareholder votes may not be informative about shareholder preferences when shareholders ignore their
own information and vote strategically conditioning on the chance of casting a pivotal vote
cost of analyzing executive pay at thousands of firms would lead shareholders to outsource
voting decisions to proxy advisors who in turn would minimize their own costs by promoting
one-size-fits-all compensation practices which would hurt firm value Finally SOP opponents
noted that shareholders had already tools to express their views on compensation matters such as
the ability to submit shareholder proposals on compensation and withhold votes from directors
responsible for compensation packages Supporters of SOP argued that enhanced shareholder
voice (as formalized in a SOP vote) and reputation concerns would help boards overcome
psychological barriers to negotiating with CEOs on behalf of shareholders resulting in more
efficient compensation contracts and more dialogue between boards and shareholders (Bebchuk
2007) They also pointed out to growing evidence of shareholdersrsquo sophistication in casting
informed votes Finally they argued that SOP would be more effective than shareholder
proposals (limited to a single issue) and less disruptive than a confrontational vote against an
otherwise valuable director More fundamentally critics and supporters of SOP disagreed on
whether existing compensation contracts were the result of an efficient labor market for talent or
instead the expression of management power over captive boards
Finally in 2010 a SOP provision found its way in the Dodd-Frank Act Ironically SOP a
provision designed to provide more alignment between shareholder and management was
adopted as part of a financial reform package aimed at deterring lsquoexcessiversquo risk taking
(excessive from the point of view of social welfare) which may be the result of the alignment of
interests between shareholders and management
As of today non-binding SOP votes (on the compensation report) have been mandated in
Australia Belgium Denmark Israel Italy Portugal while Netherlands Sweden Norway South
Africa and Switzerland adopted some version of a binding SOP vote (on the future compensation
policy) which is now also being considered by the European Union7
When putting SOP into its historical context it is also important to appreciate that SOP has come
to epitomize a broader movement toward greater shareholder democracy beyond the issue of
executive pay For example during the same period shareholder activists have long
(unsuccessfully) lobbied for a proxy access rule that would have made it easier for shareholders
to oust corporate directors and nominate their own candidates Hence many observers often
view the successes and failures of SOP as speaking to the potential effects of other reforms
aimed at increasing shareholder power making the academic research on SOP all more relevant
II The Effect of Say on Pay on Executive Compensation
A Effect of SOP votes on level and composition of executive pay
As discussed earlier United Kingdom was the first country to adopt SOP in 2002 for publicly
traded firms Ferri and Maber (2013) report that in most cases shareholders voted in favor of
compensation plans Failed SOP votes (ie greater than 50 of votes against) were rare (2 of
the sample) though highly publicized in the press However about one-fourth of the sample
firms received more than 20 of votes against
To capture the overall effect of SOP on CEO pay Ferri and Maber (2013) examine the
sensitivity of CEO pay to its economic determinants over the 2000-2005 period (ie before and
after the introduction of SOP regulation) and document a significant increase in the sensitivity of
7 Some countries have already modified their first SOP legislation In 2013 the UK added to the non-binding
backward-looking SOP vote on the current compensation report a binding forward-looking vote on the proposed
compensation policy In 2012 Australia modified the non-binding SOP vote to introduce the so-called ldquotwo-strikesrdquo
rule Under the rule if 25 of shareholders vote against a companyrsquos remuneration report at two consecutive annual
general meetings the entire board may have to stand for re-election within three months For more details about the
international development of SOP see Thomas and Van der Elst (2013)
CEO pay to poor performance They also find that this increase does not occur for a subset of
UK firms exempted from the SOP regulation (firms traded on the Alternative Investment Market
a sub-market of the London Stock Exchange with a more flexible regulatory system) suggesting
that the result reflects the impact of SOP rather than a general trend affecting all firms Besides
Ferri and Maber (2013) find that the increase is more pronounced in firms experiencing high
voting dissent and firms with high abnormal CEO pay before the adoption of SOP consistent
with a causal impact of SOP In contrast they fail to find any effect of SOP on the level of pay
Correa and Lel (2013) examine the effect of SOP laws on CEO pay using a large cross-country
sample of about 103000 firm-year observations from 39 countries including 12 countries that
adopted some (advisory or binding) version of SOP They essentially compare post-SOP CEO
pay at firms of countries adopting SOP to a control sample which includes all non-SOP
observations (that is pre-SOP CEO pay in countries eventually adopting SOP as well as CEO
pay in countries never adopting SOP) They find that firms in a post-SOP regime (i) exhibit
lower CEO pay levels (but only in countries adopting an advisory SOP vote) a finding driven by
a relative decline in equity awards and limited to CEOs and not the other top executives (ii)
higher pay-performance sensitivity (the authors do not look at positive and negative performance
separately) The current version of the paper does not examine whether these findings are more
pronounced in (or driven by) firms with excess CEO pay before the adoption of SOP and in
firms experiencing adverse SOP votes
The drawback of the research design in Correa and Lel (2013) is that it is not clear whether the
results reflect time-series changes within countries adopting SOP (pre vs post) or difference
between firms in SOP and non-SOP countries For example further tests in the paper suggest
that the lower CEO pay levels for post-SOP observations are not driven by a pay decrease
relative to the pre-SOP period (in fact country-by-country analyses indicate no change in CEO
pay levels after the adoption of SOP) but by the comparison with non-SOP countries Also while
the authors employ reasonable econometric solutions to deal with this issue some concerns
remains as to whether non-SOP countries can be a good benchmark for what would have
happened in SOP countries had SOP not been adopted (in other words if only countries with
problematic CEO practices adopt SOP it is possible that changes in CEO pay reflect factors
leading to the adoption of SOP rather than the effect of SOP laws per se) Notwithstanding these
concerns these findings are intriguing and future versions of this recent study have the potential
to provide more conclusive evidence on the effect of SOP
While in the US SOP was mandated by the Dodd-Frank Act its implementation was left to the
Securities and Exchange Commission (SEC) The final rule issued in January 2011 exempted
lsquosmallrsquo firms (ie firms with a public float below $75 million) from the SOP provision for two
years Iliev and Vitanova (2013) exploit this setting to estimate the effect of SOP on CEO
compensation In particularly using a regression discontinuity design they compare changes in
CEO pay of firms just above the SEC-imposed threshold to changes in CEO pay of firms just
below the threshold (but otherwise quite similar) They find no differences in terms of level and
composition of CEO pay and in terms of golden parachutes concluding that SOP had no impact
on CEO pay While the identification strategy chosen by the authors has the benefit of reducing
endogeneity issues it comes with the price of focusing only on fairly small firms (those around
the $75 million public float threshold) where CEO pay may not be a problem in the first place
(in which case the lack of impact of SOP would not be surprising but would not speak to its
impact at large firms traditionally the target of compensation-related criticism)
Cuntildeat Gine and Guadalupe (2013) also use a regression discontinuity design but in a different
setting In particular they analyze nonbinding shareholder proposals to adopt SOP (250
proposals between 2006 and 2010) Comparing the changes in CEO pay for firms voluntarily
adopting SOP in response to the proposals to firms not adopting SOP would be difficult since
the adoption decision is endogenous Similarly it would be difficult to compare changes in CEO
pay between firms where the proposal is passed (and thus significantly more likely to be
adopted) and firms where the proposal fails to be approved since the voting outcome is
endogenous as well
Hence Cuntildeat Gine and Guadalupe (2013) use instead a regression discontinuity design that
compares firms where the proposals receive slightly more than 50 of the votes (the threshold
for approval) to otherwise similar firms where the proposals receive slightly less than 50
Using this approach they conclude that the passing and adoption of SOP proposals is not
associated with changes in the level of CEO pay nor its composition and structure (mix of cash
and equity equity incentives level of perks and deferred pay) In some sense this finding
complements the one in Iliev and Vitanova (2013) in that it is based on large firms (shareholder
proposals are typically submitted at large SampP 500 firms) However both Cai and Walkling
(2011) and Cuntildeat Gine and Guadalupe (2013) show that firms targeted by SOP proposals do not
exhibit excess CEO pay or weaker governance8 Hence the lack of an effect at these firms may
8 Activists submitting SOP proposals chose to focus on a broad sample of large firms rather than target only firms
with problematic CEO pay practices to obtain visibility and show to policy-makers that investorrsquos support for SOP
was not limited to problematic firms (Ferri and Weber 2009)
not be surprising More generally the findings cannot be necessarily generalized to firms away
from the 50 threshold or to firms not targeted by SOP shareholder proposals
B Effect of SOP votes on compensation practices
Analyses of the effect of SOP on level and composition of CEO pay may not capture a number
of other changes to compensation contracts that may be induced by SOP votes but are not
reflected in measure of CEO pay For example the introduction of performance-based vesting
provisions is typically not reflected in the estimates of fair value of equity grants used by
researchers in measuring CEO pay Many other changes (eg terms of severance packages) will
only be reflected in measures of CEO pay contingent upon the occurrence of certain events
Hence it is important to complement the evidence in section IIA with an analysis of firmsrsquo
disclosures of changes to observable provisions of compensation contracts Another benefit of
this approach is that it captures the changes that boards explicitly present as a response to SOP
votes and thus it acts as a reality check on regression-based inferences For example a
regression-based finding of say a decrease in CEO pay level would be more credible if
supported by evidence that firms disclose a decision to reduce target levels of CEO pay in
response to an adverse SOP vote (an action that presumably they have an incentive to disclose to
get rewarded by shareholders in subsequent votes)
Two studies collect data on specific changes made to compensation contracts explicitly in
response to SOP votes in the UK and the US using firmsrsquo disclosures in their proxy statements
Ferri and Maber (2013) examine the compensation reports of a sample of UK firms before and
after the first SOP vote and find that firms experiencing higher voting dissent were significantly
more likely to remove compensation provisions criticized by investors as lsquorewards for failurersquo
relative to a matched sample of firms experiencing lower dissent For example they report that
among firms with long notice periods (implying larger severance payments ) the percentage of
high dissent firms that shortened them after the vote (800) was significantly higher than before
the vote (200) and also significantly higher than among low dissent firms after the vote
(333) They report similar findings for another controversial practice namely the presence of
retesting provisions in the performance-based vesting conditions of equity grants Most firms
indicate that these changes were the result of consultations with their major institutional
investors Firmsrsquo responsiveness to SOP votes was lsquorewardedrsquo with a substantial increase in
favorable SOP votes at the subsequent annual meeting Ferri and Maber (2013) also find that
many firms experiencing low voting dissent at the first SOP vote had removed these practices
before the vote9 highlighting the importance of accounting for ex ante effects when assessing the
impact of a new regulation
Ertimur Ferri and Oesch (2013) examine the effect of SOP on compensation practices in the US
in 2011 (first proxy season under mandatory SOP) Their key findings are generally similar to
the evidence from the UK (i) failed SOP votes are rare (about 2 of the sample) though cases
of substantial dissent are more frequent (ii) more than 55 of the firms experiencing significant
voting dissent respond by making material changes to their compensation plan during the
subsequent year (eg introduction of performance-based vesting conditions in equity grants use
of tougher performance targets removal of perks and tax gross-ups removal of controversial
provisions from severance contracts etc) usually in consultation with major institutional
9 In their sample 700 of low dissent firms shortened their notice periods during the year before the SOP vote
investors (iii) firms making changes to their compensation plans receive greater voting support
at the following SOP vote
Ertimur Ferri and Oesch (2013) also highlight the strong influence of the recommendations
released by proxy advisors (particularly Institutional Shareholder Services ISS) with a negative
recommendation being associated with 25 more votes against the compensation plan Further
evidence of the significant influence of ISS is the striking discontinuity in the relation between
the extent of SOP voting dissent and firmsrsquo responsiveness to the vote the percentage of firms
making compensation changes in response to SOP votes jumps from 32 to 72 around the
30 voting dissent threshold Why After the 2011 proxy season ISS had indicated that firms
failing to ldquoadequatelyrdquo respond to SOP voting dissent above 30 would receive a negative
recommendation in 2012 on the SOP proposal and on the election of compensation committee
members
Finally similar to the UK experience there is also anecdotal evidence of analogous
compensation changes being made by US firms ahead of the SOP vote to avoid a negative
proxy advisor recommendation and an adverse shareholder vote (Thomas Palmiter and Cotter
2013 Larcker McCall and Ormazabal 2013)
III The Effect of Say on Pay on Firm Value
In this section we examine the empirical studies on the overall effect of SOP on firm value
A Event studies around the adoption of Say on Pay regulations
As mentioned in the Introduction on April 20 2007 the House of Representatives passed a SOP
Bill by a 2-1 margin On the same day then-Senator Barack Obama introduced a companion
bill (S1181) in the Senate (which was then put on hold by the Senate Banking Committee)
While the SOP Billrsquos approval was expected (Democrats were in control of the House and
supported the SOP Bill) the 2-1 margin was unexpected suggesting some support for SOP
among Republicans
Cai and Walkling (2011) examine the market reaction to the Housersquos approval of the SOP Bill
for a sample of firms in the SampP 1500 index and document a positive reaction for firms more
likely to benefit from greater shareholder voice over executive pay namely firms with high
abnormal CEO cash pay (defined as the difference between actual CEO cash pay and the level
predicted based on known economic determinants such a size performance industry) firms
with low pay-for-performance sensitivity firms with a history of shareholders willing to vote
against compensation-related management proposals and firms with a history of responsiveness
to shareholder pressure on compensation issues In other words they find a positive reaction in
firms where the compensation problems are more severe and a say on pay vote is more likely to
trigger a response (eg more shareholders willing to vote against management and greater firmsrsquo
propensity to respond to an adverse vote) Note however that they do not find a significant
impact for firms with high abnormal equity and total CEO pay
Larcker Ormazabal and Taylor (2011) also examine the market reaction to the Housersquos approval
of the SOP Bill (in addition to other regulatory events related to executive pay and other
governance provisions) Similar to Cai and Wakling (2011) they fail to find a significant impact
for firms with high abnormal total CEO pay (they do not examine the price reaction for firms
with abnormal cash CEO pay or firms with lower pay-performance sensitivity)
A concern with both studies is that the Housersquos approval of a bill does not guarantee passage in
the Senate and approval by the White House In fact the press at the time reported that the
prospects of the bill in the Senate were uncertain (New York Times 2007) and the Bush White
House openly opposed the Bill (Associated Press 2007) Hence it is not clear the extent to
which this event increased the likelihood of a say on pay legislation Arguably as noted in
Section I the only lsquoeventrsquo that substantially increased the likelihood of say on pay legislation
was the financial crisis and the related perception of compensation abuses (eg outrage over AIG
bonuses)
Ferri and Maber (2013) argue that the announcement of the submission of SOP regulation to the
Parliament in the United Kingdom in June 2002 offers a more powerful setting for an event
study since it was largely unexpected and increased substantially the probability of SOP
adoption (Parliamentrsquos approval was virtually guaranteed) Using this event they document
positive abnormal returns for firms with excess CEO pay combined with poor performance and
for firms with controversial pay practices that weaken the penalties for poor performance (those
practices were often removed in response to SOP votes as discussed earlier in Section IIB)
They interpret their findings as consistent with shareholders perceiving SOP as a value
enhancing monitoring mechanism for firms with low pay-to-poor-performance sensitivity
Finally a recent study by Iliev and Vitanova (2013) examines the market reaction around the
announcement of the SEC final SOP rule that on January 25 2011 exempted small firms (ie
firms with a public float below $75 million) from adopting SOP for two years (press reports at
the time noted that the exemption could become permanent) An appealing feature of this setting
is that the SEC decision was an unexpected reversal from the rule proposed in October 2010 rule
under which no publicly traded firm would be exempted
Iliev and Vitanova (2013) compare the market reaction to this announcement for firms just above
the SEC-imposed threshold and thus subject to the SOP rule and firms just below the threshold
(but otherwise quite similar) and thus exempted from SOP While they find not significant
abnormal returns for either group the abnormal returns for exempted firms are significantly
more negative a finding that the authors interpret as evidence of a positive value effect of
mandatory SOP While the setting is an interesting one some concerns remain as exempted
firms are fairly small and the magnitude of the executive pay problem (if any) is unlikely to be
large enough to explain the differential market reaction Also the same study finds no effect on
CEO pay level and composition (see Section IIA) Hence the reason for the differential stock
price reaction remains unclear It would be helpful to know if the event study result is driven by
(or stronger for) exempted firms with excessive pay or problematic pay practices
B Event studies around Shareholder Proposals to Adopt Say on Pay
Starting in 2006 shareholder activists led by union pension funds submitted shareholder
proposal to adopt SOP at hundreds of US firms in an attempt to induce voluntary or mandatory
widespread adoption of SOP
Cai and Walkling (2011) examine the market reaction to proxy filings and annual meetings of
113 firms targeted by nonbinding shareholder proposals to adopt SOP between 2006 and 2008
On average they find insignificant returns around both the submission of the proposal (proxy
filing date) and the vote on the proposal (annual meeting date) But they also find that when the
proposals are filed by union pension funds the abnormal returns (still insignificant) are more
negative than when filed by other activists and that when the proposal is defeated the abnormal
returns (while insignificant) are more positive than when the proposal is passed (it is not clear
whether this result is driven by union-sponsored proposals) They interpret these findings as
evidence that the market views SOP proposals filed by union pension funds are driven by special
interests rather than value maximization10
However caution is required in interpreting this
evidence Virtually all activists filing SOP proposals were coordinated by a group of investors
(mostly but not only union pension funds) who made available to other activists a template for
SOP proposals and a list of target firms (Ferri and Weber 2009) Hence the distinction between
union and non-union proponents in the context of SOP proposals may be only apparent More
importantly the list of target firms was publicly available months before the proxy filing dates
so it is not clear whether the proxy statements contained any new information Similarly the
voting outcome of the SOP proposals may largely be anticipated (based on the composition of
institutional owners proxy advisorsrsquo recommendations etc) Finally the analyses do not control
for other (potentially new) information contained in proxy statements (eg executive
compensation report other items up for a vote at the annual meeting) or other events occurring at
the annual meeting (eg shareholder votes on other items)
Cuntildeat Gine and Guadalupe (2013) also examine the market reaction to the voting outcome of
nonbinding shareholder proposals to adopt SOP but to alleviate these concerns they employ a
lsquofuzzyrsquo regression discontinuity design (RDD) essentially comparing the stock price reaction to
SOP proposals that pass by a small margin to the reaction to SOP proposals that fail by a small
10
Consistent with this interpretation the authors also find that SOP proposals generally targeted larger firms rather
than firms with excessive pay or lower pay-performance sensitivity However as noted by Ferri and Sandino (2009)
activists typically submit shareholder proposals aimed at promoting policy reforms at a broad sample of large firms
rather than the firms that would most benefit from the proposals because they believe that showing widespread
support for the proposal across all types of firms enhances its credibility with policy makers
margin (Cuntildeat Gine and Guadalupe 2012 Ertimur Ferri and Oesch forthcoming) The underlying idea
is that firms around the threshold are likely to have similar characteristics (indeed they do as the
authors show) but differ in the likelihood of implementation which is typically much higher for
proposals passing the threshold (Ertimur Ferri and Stubben 2010 Thomas and Cotter 2007)
Indeed the authors show that the likelihood of subsequently adopting SOP is 48 higher when
the SOP proposal passes by a small margin relatively to when it fails by a small margin Besides
because in these close-call situations the voting outcome is uncertain its resolution (pass or fail)
is likely to convey new information about the likelihood of SOP adoption making this setting
suitable to an event study11
Using this RDD approach Cuntildeat Gine and Guadalupe (2013) find that on the day of the vote a
SOP proposal that passes by a small margin yields an abnormal return of 24 relative to one
that fails (after controlling for other proposals voted upon at the same meeting) and estimate the
lsquofullrsquo value of SOP at 46 (after taking into account the increase in the probability of SOP
implementation)12
Aside from the issues discussed in Section IA (ie generalizability to other
firms) this estimate seems too large to reflect the present value of future reductions in excess
CEO pay (as noted by the authors their estimate of the value of SOP is equivalent to the
estimated value of removing two anti-takeover provision based on Cuntildeat Gine and Guadalupe
2012) particularly because (i) the authors find no evidence of subsequent changes in levels and
composition of CEO pay for firms implementing SOP (see Section IIA) an (ii) the sample firms
11
Consistent with this observation the authors find no market reaction to the voting outcome of SOP shareholder
proposals away from the threshold
12 Since the outcome of the vote is not binding the 24 market reaction only reflects the expected increase in the
probability of SOP adoption and thus understates the value of the SOP provision
do not appear to be characterized by excess CEO pay However the authors also find significant
improvements in subsequent operating performance and efficiency as a result of passing SOP
proposals and thus conjecture that the SOP vote is viewed as a tool to express a vote of
confidence in management performance more than a way to change compensation practices and
that the large positive market reaction reflects expected performance improvements under this
tighter monitoring regime While this is an interesting idea it remains unclear what additional
ldquoteethrdquo a SOP vote may provide over existing (non-compensation related) monitoring
mechanisms shareholders can express (and do often express) their dissatisfaction with
management performance when voting on director elections (Cai Garner and Walkling 2009)
C Event studies around compensation changes induced by SOP
Two studies examine announcements of changes to compensation plans related to SOP Larcker
McCall and Ormazabal (2013) examine the market reaction to (non-contaminated) compensation
changes disclosed in 8-K filings by Russell 3000 firms in the US during the year before the first
SOP vote They find that changes that appear to be made to avoid a negative proxy advisorrsquos
recommendation and thus a negative SOP vote are associated with a negative abnormal return
of -044 whereas other compensation changes are not associated with a significant stock price
reaction Ertimur Ferri and Oesch (2013) perform a similar analysis but focusing on
compensation changes made after the first SOP vote (and explicitly in response to the vote) by
firms receiving a negative recommendation and a high voting dissent in 2011 They fail to find
significant abnormal returns even for the subset of compensation changes that resulted in a
positive recommendation and low dissent in 2012 (and thus were presumably perceived to be
adequate and material by proxy advisors and voting shareholders)
D Other approaches effect of SOP on Tobinrsquos Q
In their cross-country study Correa and Lel (2013) also examine the effect of SOP laws on
Tobinrsquos Q and report a 36 increase in firm value following the adoption of the SOP laws
(more precisely a 36 higher firm value for post-SOP observations relative to a control sample
that pools together pre-SOP and non-SOP observations) They acknowledge that this increase in
firm value is too large to be justified by the relative decrease in CEO pay that they document and
suggest (but do not test) that it may reflect better alignment of pay and performance13
A key
concern is that higher valuation in countries with SOP laws may reflect other governance
changes introduced at the same time While the study controls for other compensation-related
laws there may be other non-compensation regulations introduced with SOP and with a
potentially larger impact on firm value
IV What have we learned
Adoption of SOP in the US has been long advocated by those who contend that executive pay is
a manifestation of rather than a solution to the agency problem and by those who favor great
shareholder involvement in corporate decisions A growing body of research is starting to
investigate the effect of SOP on executive pay and firm value It is premature to draw definitive
conclusions also because these studies differ in methodologies and settings but three points
seem to emerge First there is robust evidence that boards respond to SOP votes when
13
They also find that the firm value increase is higher when the relative decrease in CEO pay results in a lower pay
differential between CEO and other top managers (CEO pay slice) implying that a source of value creation may be
the reduced pay inequality among the top management team But this seems unlikely to explain the change in
Tobinrsquos Q since the increase occurs for both firms with advisory SOP laws and firms with binding SOP laws and
there is no change in CEO pay slice in firms subject to binding SOP laws
shareholders use it ldquovoicerdquo is heardrdquo Both in the US and UK studies show that not only firms
failing to win the SOP vote but also firms facing substantial dissent (20-30 of the votes
against) make changes to their compensation plans in consultation with institutional investors
and proxy advisors Second in both the US and UK institutional investors appear to use the
power of SOP votes to pressure firms into strengthening the link (or the perceived link) between
pay and performance (particularly on the downside) but not to pressure firms to reduce target
levels of pay suggesting a reluctance of institutional investors to lsquoregulatersquo executive pay Third
and consistent with the above most studies examining the aggregate effect of SOP on CEO pay
find some increase in pay-performance sensitivity but not effect on the level and growth of CEO
pay While some observers may interpret the lack of an effect on CEO pay levels as evidence of
the failure of SOP one should note that SOP per se is a neutral tool Its impact depends on what
investors choose to lsquosay on payrsquo Fourth based on my interpretation of the existing research to
date there is little evidence of an effect of SOP on firm value The studies documenting a
positive price impact appear subject to alternative interpretations or not plausible in their
findings (eg they fail to find an effect on CEO pay or the effect is too small to justify the
documented price impact hence it remains unclear what the source of value creation is) Also
there seems to be no stock price reaction to the SOP-induced compensation changes Finally
given the nature of these SOP-induced changes while some of them may be beneficial it seems
hard to believe that they will have a statistically detectable impact on firm value except perhaps
in a handful of firms where the compensation plan is subject to a complete overhaul
Perhaps one explanation for the weak impact of SOP is that executive pay problems were
overstated or that by the time SOP was introduced (more than a decade after the Enron-type
scandals that led to calls for greater shareholder voice) they had been already addressed via
other mechanisms (hedge fund activism monitoring by institutional investors vote-no
campaigns against compensation committee members SEC-mandated pay disclosures)
Overall though based on the evidence to date it does not appear that SOP had a major impact14
Was the adoption of SOP an lsquooptimalrsquo choice from a social welfare point of view This is a
difficult question to answer given the challenges of identifying and measuring all the potential
costs and benefits associated with SOP (likely to change over time and differ across countries)
Perhaps the most positive view of SOP is that its introduction prevented the adoption of other
more radical and intrusive regulatory measures (eg CEO pay caps) In that sense it may be
viewed as an lsquooptimalrsquo answer (ie the lesser of two evils) to the political pressure to reform
executive pay during the financial crisis
14
This interpretation of the evidence is also consistent with the prediction of analytical studies Ozbas and
Matsusaka (2013) model the benefits and costs of shareholder empowerment distinguishing between the right to
approve and the right to propose They show that permitting shareholders to propose directors or policies can cause
value-maximizing managers to take value-reducing actions to accommodate activist investors with non-value-
maximizing goals As for the right to approve it is weakly beneficial that is approval rights (such as a SOP vote)
are generally beneficial for shareholders in that they limit the managerrsquos ability to pursue private benefits at
shareholder expense but they have minimal impact on managerial actions and firm value (because the manager in
effect can threaten shareholders with an undesirable status quo if they do not approve the managerrsquos proposed
action)
References
Associated Press 2007 Administration opposes say on pay bill April 19
Bainbridge S 2008 Remarks on say on pay An unjustified incursion on director authority
UCLA School of Law Research Paper No 08-06
Bebchuk L 2005 The Case for Increasing Shareholder Power Harvard Law Review 118833-
917
Bebchuk L 2007 Written testimony submitted before the Committee on Financial Services
United States House of Representatives Hearing on Empowering Shareholders on Executive
Compensation March 8
Bebchuk LA A Cohen and A Ferrell 2009 What Matters in Corporate Governance Review
of Financial Studies 22783-827
Cai J J Garner and R Walkling 2009 Electing Directors Journal of Finance 642387-2419
Cai J and R Walkling 2011 Shareholdersrsquo Say on Pay Does it Create Value Journal of
Financial and Quantitative Analysis 46299ndash339
Coates IV JC (2009) Testimony before the Subcommittee on Securities Insurance and
Investment of the Committee on Banking Housing and Urban Affairs United States Senate
July 29 2009 httppapersssrncomsol3paperscfmabstract_id=1475355
Correa R and U Lel 2013 Say On Pay Laws Executive Compensation CEO Pay Slice and
Firm Value Around the World Federal Reserve Board Working Paper
Cuntildeat V M Gine and M Guadalupe 2012 The Vote is Cast The Effect of Corporate
Governance on Shareholder Value Journal of Finance 671943-1977
Cuntildeat V M Gine and M Guadalupe 2013 Say Pays Shareholder Voice and Firm
Performance ECGI - Finance Working Paper No 373
Del Guercio D L Seery and T Woidtke 2008 Do Boards Pay Attention When Institutional
Investors lsquoJust Vote Norsquo Journal of Financial Economics 9084-103
Ertimur Y F Ferri and S Stubben 2010 Board of Directorslsquo Responsiveness to Shareholders
Evidence from Shareholder Proposals Journal of Corporate Finance 1653-72
Ertimur Y F Ferri and V Muslu 2011 Shareholder Activism and CEO Pay Review of
Financial Studies 24535ndash592
Ertimur Y F Ferri and D Oesch 2013 Shareholder Votes and Proxy Advisors Evidence from
Say on Pay Journal of Accounting Research 51951-996
Ertimur Y F Ferri and D Oesch forthcoming Does the Director Election System Matter
Evidence from Majority Voting Review of Accounting Studies forthcoming
Ferri F 2012 Low-Cost Shareholder Activism A Review of the Evidence Research
Handbook on the Economics of Corporate Law ed CA Hill and BH McDonnell
Edward Elgar Publishing
Ferri F and D Maber 2013 Say on Pay Votes and CEO Compensation Evidence from the UK
Review of Finance 17527-563
Ferri F and D Oesch 2013 Management Influence on Investors Evidence from Shareholder
Votes on the Frequency of Say on Pay Columbia Business School Research Paper No 13-17
Ferri F and T Sandino 2009 The impact of shareholder activism on financial reporting and
compensation The case of employee stock options expensing The Accounting Review 84433ndash
466
Ferri F and J Weber 2009 AFSCME vs Mozilohellipand Say on Pay for All (A) Harvard
Business School Case 109-009
Gompers P J Ishii and A Metrick 2003 Corporate Governance and Equity Prices Quarterly
Journal of Economics 118107-155
Gordon J 2009 Say on Paylsquo Cautionary Notes on the UK Experience and the Case for
Shareholders Opt-In Harvard Journal on Legislation 46323-368
Iliev P and S Vitanova 2013 The effect of the Say on Pay in the US Pennsylvania State
University Working Paper
Kaplan S 2007 Testimony of Steven N Kaplan on Empowering Shareholders on Executive
Compensation and HR 1257 the Shareholder Vote on Executive Compensation Act before the
Committee on Financial Services United States House of Representatives March 8
Larcker D A McCall and G Ormazabal 2013 The Economic Consequences of Proxy
Advisor Say-on-Pay Voting Policies Working Paper Stanford University
Larcker D G Ormazabal and D Taylor 2011 The market reaction to corporate governance
regulation Journal of Financial Economics 101431ndash448
Levit D and N Malenko 2011 Non-binding voting for shareholder proposals Journal of
Finance 661579ndash1614
Mishel L and N Sabadish 2012 How executive compensation and financial-sector pay have
fueled income inequality Issue Brief 331 Economic Policy Institute
Murphy K 2012 The Politics of Pay A Legislative History of Executive Compensation in
Jennifer Hill and Randall Thomas eds Research Handbook on Executive Pay Edward Elgar
Publishers
New York Times 2007 House votes to give investors say on executive pay April 21
Norris F 2004 Corporate democracy and the power to embarrass New York Times March 4
Ozbas O and J Matsusaka2013 Managerial Accommodation Proxy Access and the Cost of
Shareholder Empowerment University of Southern California CLEO Research Paper Series No
C12-1
Thomas R and J Cotter 2007 Shareholder proposals in the new millennium Shareholder
support board response and market reaction Journal of Corporate Finance 13 368ndash391
Thomas R A Palmiter and J Cotter 2012 Dodd-Franks Say on Pay Will it Lead to a Greater
Role for Shareholders in Corporate Governance Cornell Law Review 971213-1266
Thomas R and C Van der Elst 2013 The International Scope of Say on Pay ECGI Law
Working Paper No227
failed say on pay vote at Glaxo SmithKline (with 507 of the votes cast against approval of the
remuneration report) made headlines around the world Shareholders had objected to an
estimated pound22 million severance arrangement for the CEO (reflecting a two-year notice period)
lack of challenging performance targets and the presence of a retesting provision in the stock
option plan4 Glaxo SmithKlinersquos board responded to the vote by launching an extensive
consultation process with shareholders and adjusting the compensation package accordingly
(Ferri and Maber 2013)
In the United States the American Federation of State County and Municipal Employees a
union pension fund took notice of the UK experience and rallied other activist investors to
submit shareholder proposals requesting the adoption of SOP (Ferri and Weber 2009) Between
2006 and 2010 SOP proposals were submitted and voted upon at hundreds of US firms
averaging more than 43 votes in favor and often winning a majority vote with increasing
success over the years (Cuntildeat Gine and Guadalupe 2013) High-profile executive pay scandals
(eg option backdating the large severance package awarded to Nardelli at Home Depot) helped
activists in making their case for more ldquosay on payrdquo among investors and policy makers While
submitted at individual companies the objective of the SOP proposals was to induce widespread
adoption of SOP either via voluntary adoption (which turned out to be rare) a listing standard
or regulatory intervention Policy makers took notice and on April 20 2007 the House of
Representatives passed a bill requiring a nonbinding annual shareholder vote on executive
4 Retesting provisions in the performance-based vesting conditions of equity grants allow to reevaluate in
subsequent years performance targets not achieved during the initial measurement period rather than allowing the
equity grant to lapse As such they were criticized in the UK as a form of lsquoreward for failurersquo
compensation (hereafter SOP Bill)5 On the same day then-Senator Barack Obama introduced a
companion bill in the Senate (S1181)
But it was the financial crisis of 2007-2008 to accelerate the path toward mandatory
adoption of SOP Growing income inequality public outrage over Wall Street excesses (eg the
large bonuses paid at AIG) and banksrsquo bailouts and the perception that executive pay played a
role in inducing excessive risk-taking pressured policy-makers to take action with respect to
executive pay Many of the legislative proposals discussed in the House and Senate in 2008 and
2009 contained a SOP provision (see Table 1 in Larcker Ormazabal and Taylor 2011) Also
holding a SOP vote was a mandatory condition for firms to receive funds under the Troubled
Asset Relief Program (TARP) During the 2008 campaign both Presidential candidates
expressed support for SOP As these events unfolded the debate on merits and drawbacks of
SOP heated up Critics of SOP argued that at best SOP votes would be ignored (because
nonbinding) and at worst would cause directors to pander to shareholders with special interests
or lacking the required expertise and sophistication ultimately resulting in the adoption of
suboptimal pay practices (Kaplan 2007 Bainbridge 2008)6 They also cautioned that the high
5 House Bill 1257 Shareholder Vote on Executive Compensation Act
6 Analytical studies have also raised questions on the informativeness of SOP votes Ozbas and Matsusaka (2013)
note that SOP votes only reveal shareholder preferences for the current compensation plan compared to the
alternative plan that will be adopted if the current plan is rejected but such alternative plan is unclear in practice
casting doubts on whether the shareholder votes on SOP may be truly informative Levit and Malenko (2011) further
suggest that shareholder votes may not be informative about shareholder preferences when shareholders ignore their
own information and vote strategically conditioning on the chance of casting a pivotal vote
cost of analyzing executive pay at thousands of firms would lead shareholders to outsource
voting decisions to proxy advisors who in turn would minimize their own costs by promoting
one-size-fits-all compensation practices which would hurt firm value Finally SOP opponents
noted that shareholders had already tools to express their views on compensation matters such as
the ability to submit shareholder proposals on compensation and withhold votes from directors
responsible for compensation packages Supporters of SOP argued that enhanced shareholder
voice (as formalized in a SOP vote) and reputation concerns would help boards overcome
psychological barriers to negotiating with CEOs on behalf of shareholders resulting in more
efficient compensation contracts and more dialogue between boards and shareholders (Bebchuk
2007) They also pointed out to growing evidence of shareholdersrsquo sophistication in casting
informed votes Finally they argued that SOP would be more effective than shareholder
proposals (limited to a single issue) and less disruptive than a confrontational vote against an
otherwise valuable director More fundamentally critics and supporters of SOP disagreed on
whether existing compensation contracts were the result of an efficient labor market for talent or
instead the expression of management power over captive boards
Finally in 2010 a SOP provision found its way in the Dodd-Frank Act Ironically SOP a
provision designed to provide more alignment between shareholder and management was
adopted as part of a financial reform package aimed at deterring lsquoexcessiversquo risk taking
(excessive from the point of view of social welfare) which may be the result of the alignment of
interests between shareholders and management
As of today non-binding SOP votes (on the compensation report) have been mandated in
Australia Belgium Denmark Israel Italy Portugal while Netherlands Sweden Norway South
Africa and Switzerland adopted some version of a binding SOP vote (on the future compensation
policy) which is now also being considered by the European Union7
When putting SOP into its historical context it is also important to appreciate that SOP has come
to epitomize a broader movement toward greater shareholder democracy beyond the issue of
executive pay For example during the same period shareholder activists have long
(unsuccessfully) lobbied for a proxy access rule that would have made it easier for shareholders
to oust corporate directors and nominate their own candidates Hence many observers often
view the successes and failures of SOP as speaking to the potential effects of other reforms
aimed at increasing shareholder power making the academic research on SOP all more relevant
II The Effect of Say on Pay on Executive Compensation
A Effect of SOP votes on level and composition of executive pay
As discussed earlier United Kingdom was the first country to adopt SOP in 2002 for publicly
traded firms Ferri and Maber (2013) report that in most cases shareholders voted in favor of
compensation plans Failed SOP votes (ie greater than 50 of votes against) were rare (2 of
the sample) though highly publicized in the press However about one-fourth of the sample
firms received more than 20 of votes against
To capture the overall effect of SOP on CEO pay Ferri and Maber (2013) examine the
sensitivity of CEO pay to its economic determinants over the 2000-2005 period (ie before and
after the introduction of SOP regulation) and document a significant increase in the sensitivity of
7 Some countries have already modified their first SOP legislation In 2013 the UK added to the non-binding
backward-looking SOP vote on the current compensation report a binding forward-looking vote on the proposed
compensation policy In 2012 Australia modified the non-binding SOP vote to introduce the so-called ldquotwo-strikesrdquo
rule Under the rule if 25 of shareholders vote against a companyrsquos remuneration report at two consecutive annual
general meetings the entire board may have to stand for re-election within three months For more details about the
international development of SOP see Thomas and Van der Elst (2013)
CEO pay to poor performance They also find that this increase does not occur for a subset of
UK firms exempted from the SOP regulation (firms traded on the Alternative Investment Market
a sub-market of the London Stock Exchange with a more flexible regulatory system) suggesting
that the result reflects the impact of SOP rather than a general trend affecting all firms Besides
Ferri and Maber (2013) find that the increase is more pronounced in firms experiencing high
voting dissent and firms with high abnormal CEO pay before the adoption of SOP consistent
with a causal impact of SOP In contrast they fail to find any effect of SOP on the level of pay
Correa and Lel (2013) examine the effect of SOP laws on CEO pay using a large cross-country
sample of about 103000 firm-year observations from 39 countries including 12 countries that
adopted some (advisory or binding) version of SOP They essentially compare post-SOP CEO
pay at firms of countries adopting SOP to a control sample which includes all non-SOP
observations (that is pre-SOP CEO pay in countries eventually adopting SOP as well as CEO
pay in countries never adopting SOP) They find that firms in a post-SOP regime (i) exhibit
lower CEO pay levels (but only in countries adopting an advisory SOP vote) a finding driven by
a relative decline in equity awards and limited to CEOs and not the other top executives (ii)
higher pay-performance sensitivity (the authors do not look at positive and negative performance
separately) The current version of the paper does not examine whether these findings are more
pronounced in (or driven by) firms with excess CEO pay before the adoption of SOP and in
firms experiencing adverse SOP votes
The drawback of the research design in Correa and Lel (2013) is that it is not clear whether the
results reflect time-series changes within countries adopting SOP (pre vs post) or difference
between firms in SOP and non-SOP countries For example further tests in the paper suggest
that the lower CEO pay levels for post-SOP observations are not driven by a pay decrease
relative to the pre-SOP period (in fact country-by-country analyses indicate no change in CEO
pay levels after the adoption of SOP) but by the comparison with non-SOP countries Also while
the authors employ reasonable econometric solutions to deal with this issue some concerns
remains as to whether non-SOP countries can be a good benchmark for what would have
happened in SOP countries had SOP not been adopted (in other words if only countries with
problematic CEO practices adopt SOP it is possible that changes in CEO pay reflect factors
leading to the adoption of SOP rather than the effect of SOP laws per se) Notwithstanding these
concerns these findings are intriguing and future versions of this recent study have the potential
to provide more conclusive evidence on the effect of SOP
While in the US SOP was mandated by the Dodd-Frank Act its implementation was left to the
Securities and Exchange Commission (SEC) The final rule issued in January 2011 exempted
lsquosmallrsquo firms (ie firms with a public float below $75 million) from the SOP provision for two
years Iliev and Vitanova (2013) exploit this setting to estimate the effect of SOP on CEO
compensation In particularly using a regression discontinuity design they compare changes in
CEO pay of firms just above the SEC-imposed threshold to changes in CEO pay of firms just
below the threshold (but otherwise quite similar) They find no differences in terms of level and
composition of CEO pay and in terms of golden parachutes concluding that SOP had no impact
on CEO pay While the identification strategy chosen by the authors has the benefit of reducing
endogeneity issues it comes with the price of focusing only on fairly small firms (those around
the $75 million public float threshold) where CEO pay may not be a problem in the first place
(in which case the lack of impact of SOP would not be surprising but would not speak to its
impact at large firms traditionally the target of compensation-related criticism)
Cuntildeat Gine and Guadalupe (2013) also use a regression discontinuity design but in a different
setting In particular they analyze nonbinding shareholder proposals to adopt SOP (250
proposals between 2006 and 2010) Comparing the changes in CEO pay for firms voluntarily
adopting SOP in response to the proposals to firms not adopting SOP would be difficult since
the adoption decision is endogenous Similarly it would be difficult to compare changes in CEO
pay between firms where the proposal is passed (and thus significantly more likely to be
adopted) and firms where the proposal fails to be approved since the voting outcome is
endogenous as well
Hence Cuntildeat Gine and Guadalupe (2013) use instead a regression discontinuity design that
compares firms where the proposals receive slightly more than 50 of the votes (the threshold
for approval) to otherwise similar firms where the proposals receive slightly less than 50
Using this approach they conclude that the passing and adoption of SOP proposals is not
associated with changes in the level of CEO pay nor its composition and structure (mix of cash
and equity equity incentives level of perks and deferred pay) In some sense this finding
complements the one in Iliev and Vitanova (2013) in that it is based on large firms (shareholder
proposals are typically submitted at large SampP 500 firms) However both Cai and Walkling
(2011) and Cuntildeat Gine and Guadalupe (2013) show that firms targeted by SOP proposals do not
exhibit excess CEO pay or weaker governance8 Hence the lack of an effect at these firms may
8 Activists submitting SOP proposals chose to focus on a broad sample of large firms rather than target only firms
with problematic CEO pay practices to obtain visibility and show to policy-makers that investorrsquos support for SOP
was not limited to problematic firms (Ferri and Weber 2009)
not be surprising More generally the findings cannot be necessarily generalized to firms away
from the 50 threshold or to firms not targeted by SOP shareholder proposals
B Effect of SOP votes on compensation practices
Analyses of the effect of SOP on level and composition of CEO pay may not capture a number
of other changes to compensation contracts that may be induced by SOP votes but are not
reflected in measure of CEO pay For example the introduction of performance-based vesting
provisions is typically not reflected in the estimates of fair value of equity grants used by
researchers in measuring CEO pay Many other changes (eg terms of severance packages) will
only be reflected in measures of CEO pay contingent upon the occurrence of certain events
Hence it is important to complement the evidence in section IIA with an analysis of firmsrsquo
disclosures of changes to observable provisions of compensation contracts Another benefit of
this approach is that it captures the changes that boards explicitly present as a response to SOP
votes and thus it acts as a reality check on regression-based inferences For example a
regression-based finding of say a decrease in CEO pay level would be more credible if
supported by evidence that firms disclose a decision to reduce target levels of CEO pay in
response to an adverse SOP vote (an action that presumably they have an incentive to disclose to
get rewarded by shareholders in subsequent votes)
Two studies collect data on specific changes made to compensation contracts explicitly in
response to SOP votes in the UK and the US using firmsrsquo disclosures in their proxy statements
Ferri and Maber (2013) examine the compensation reports of a sample of UK firms before and
after the first SOP vote and find that firms experiencing higher voting dissent were significantly
more likely to remove compensation provisions criticized by investors as lsquorewards for failurersquo
relative to a matched sample of firms experiencing lower dissent For example they report that
among firms with long notice periods (implying larger severance payments ) the percentage of
high dissent firms that shortened them after the vote (800) was significantly higher than before
the vote (200) and also significantly higher than among low dissent firms after the vote
(333) They report similar findings for another controversial practice namely the presence of
retesting provisions in the performance-based vesting conditions of equity grants Most firms
indicate that these changes were the result of consultations with their major institutional
investors Firmsrsquo responsiveness to SOP votes was lsquorewardedrsquo with a substantial increase in
favorable SOP votes at the subsequent annual meeting Ferri and Maber (2013) also find that
many firms experiencing low voting dissent at the first SOP vote had removed these practices
before the vote9 highlighting the importance of accounting for ex ante effects when assessing the
impact of a new regulation
Ertimur Ferri and Oesch (2013) examine the effect of SOP on compensation practices in the US
in 2011 (first proxy season under mandatory SOP) Their key findings are generally similar to
the evidence from the UK (i) failed SOP votes are rare (about 2 of the sample) though cases
of substantial dissent are more frequent (ii) more than 55 of the firms experiencing significant
voting dissent respond by making material changes to their compensation plan during the
subsequent year (eg introduction of performance-based vesting conditions in equity grants use
of tougher performance targets removal of perks and tax gross-ups removal of controversial
provisions from severance contracts etc) usually in consultation with major institutional
9 In their sample 700 of low dissent firms shortened their notice periods during the year before the SOP vote
investors (iii) firms making changes to their compensation plans receive greater voting support
at the following SOP vote
Ertimur Ferri and Oesch (2013) also highlight the strong influence of the recommendations
released by proxy advisors (particularly Institutional Shareholder Services ISS) with a negative
recommendation being associated with 25 more votes against the compensation plan Further
evidence of the significant influence of ISS is the striking discontinuity in the relation between
the extent of SOP voting dissent and firmsrsquo responsiveness to the vote the percentage of firms
making compensation changes in response to SOP votes jumps from 32 to 72 around the
30 voting dissent threshold Why After the 2011 proxy season ISS had indicated that firms
failing to ldquoadequatelyrdquo respond to SOP voting dissent above 30 would receive a negative
recommendation in 2012 on the SOP proposal and on the election of compensation committee
members
Finally similar to the UK experience there is also anecdotal evidence of analogous
compensation changes being made by US firms ahead of the SOP vote to avoid a negative
proxy advisor recommendation and an adverse shareholder vote (Thomas Palmiter and Cotter
2013 Larcker McCall and Ormazabal 2013)
III The Effect of Say on Pay on Firm Value
In this section we examine the empirical studies on the overall effect of SOP on firm value
A Event studies around the adoption of Say on Pay regulations
As mentioned in the Introduction on April 20 2007 the House of Representatives passed a SOP
Bill by a 2-1 margin On the same day then-Senator Barack Obama introduced a companion
bill (S1181) in the Senate (which was then put on hold by the Senate Banking Committee)
While the SOP Billrsquos approval was expected (Democrats were in control of the House and
supported the SOP Bill) the 2-1 margin was unexpected suggesting some support for SOP
among Republicans
Cai and Walkling (2011) examine the market reaction to the Housersquos approval of the SOP Bill
for a sample of firms in the SampP 1500 index and document a positive reaction for firms more
likely to benefit from greater shareholder voice over executive pay namely firms with high
abnormal CEO cash pay (defined as the difference between actual CEO cash pay and the level
predicted based on known economic determinants such a size performance industry) firms
with low pay-for-performance sensitivity firms with a history of shareholders willing to vote
against compensation-related management proposals and firms with a history of responsiveness
to shareholder pressure on compensation issues In other words they find a positive reaction in
firms where the compensation problems are more severe and a say on pay vote is more likely to
trigger a response (eg more shareholders willing to vote against management and greater firmsrsquo
propensity to respond to an adverse vote) Note however that they do not find a significant
impact for firms with high abnormal equity and total CEO pay
Larcker Ormazabal and Taylor (2011) also examine the market reaction to the Housersquos approval
of the SOP Bill (in addition to other regulatory events related to executive pay and other
governance provisions) Similar to Cai and Wakling (2011) they fail to find a significant impact
for firms with high abnormal total CEO pay (they do not examine the price reaction for firms
with abnormal cash CEO pay or firms with lower pay-performance sensitivity)
A concern with both studies is that the Housersquos approval of a bill does not guarantee passage in
the Senate and approval by the White House In fact the press at the time reported that the
prospects of the bill in the Senate were uncertain (New York Times 2007) and the Bush White
House openly opposed the Bill (Associated Press 2007) Hence it is not clear the extent to
which this event increased the likelihood of a say on pay legislation Arguably as noted in
Section I the only lsquoeventrsquo that substantially increased the likelihood of say on pay legislation
was the financial crisis and the related perception of compensation abuses (eg outrage over AIG
bonuses)
Ferri and Maber (2013) argue that the announcement of the submission of SOP regulation to the
Parliament in the United Kingdom in June 2002 offers a more powerful setting for an event
study since it was largely unexpected and increased substantially the probability of SOP
adoption (Parliamentrsquos approval was virtually guaranteed) Using this event they document
positive abnormal returns for firms with excess CEO pay combined with poor performance and
for firms with controversial pay practices that weaken the penalties for poor performance (those
practices were often removed in response to SOP votes as discussed earlier in Section IIB)
They interpret their findings as consistent with shareholders perceiving SOP as a value
enhancing monitoring mechanism for firms with low pay-to-poor-performance sensitivity
Finally a recent study by Iliev and Vitanova (2013) examines the market reaction around the
announcement of the SEC final SOP rule that on January 25 2011 exempted small firms (ie
firms with a public float below $75 million) from adopting SOP for two years (press reports at
the time noted that the exemption could become permanent) An appealing feature of this setting
is that the SEC decision was an unexpected reversal from the rule proposed in October 2010 rule
under which no publicly traded firm would be exempted
Iliev and Vitanova (2013) compare the market reaction to this announcement for firms just above
the SEC-imposed threshold and thus subject to the SOP rule and firms just below the threshold
(but otherwise quite similar) and thus exempted from SOP While they find not significant
abnormal returns for either group the abnormal returns for exempted firms are significantly
more negative a finding that the authors interpret as evidence of a positive value effect of
mandatory SOP While the setting is an interesting one some concerns remain as exempted
firms are fairly small and the magnitude of the executive pay problem (if any) is unlikely to be
large enough to explain the differential market reaction Also the same study finds no effect on
CEO pay level and composition (see Section IIA) Hence the reason for the differential stock
price reaction remains unclear It would be helpful to know if the event study result is driven by
(or stronger for) exempted firms with excessive pay or problematic pay practices
B Event studies around Shareholder Proposals to Adopt Say on Pay
Starting in 2006 shareholder activists led by union pension funds submitted shareholder
proposal to adopt SOP at hundreds of US firms in an attempt to induce voluntary or mandatory
widespread adoption of SOP
Cai and Walkling (2011) examine the market reaction to proxy filings and annual meetings of
113 firms targeted by nonbinding shareholder proposals to adopt SOP between 2006 and 2008
On average they find insignificant returns around both the submission of the proposal (proxy
filing date) and the vote on the proposal (annual meeting date) But they also find that when the
proposals are filed by union pension funds the abnormal returns (still insignificant) are more
negative than when filed by other activists and that when the proposal is defeated the abnormal
returns (while insignificant) are more positive than when the proposal is passed (it is not clear
whether this result is driven by union-sponsored proposals) They interpret these findings as
evidence that the market views SOP proposals filed by union pension funds are driven by special
interests rather than value maximization10
However caution is required in interpreting this
evidence Virtually all activists filing SOP proposals were coordinated by a group of investors
(mostly but not only union pension funds) who made available to other activists a template for
SOP proposals and a list of target firms (Ferri and Weber 2009) Hence the distinction between
union and non-union proponents in the context of SOP proposals may be only apparent More
importantly the list of target firms was publicly available months before the proxy filing dates
so it is not clear whether the proxy statements contained any new information Similarly the
voting outcome of the SOP proposals may largely be anticipated (based on the composition of
institutional owners proxy advisorsrsquo recommendations etc) Finally the analyses do not control
for other (potentially new) information contained in proxy statements (eg executive
compensation report other items up for a vote at the annual meeting) or other events occurring at
the annual meeting (eg shareholder votes on other items)
Cuntildeat Gine and Guadalupe (2013) also examine the market reaction to the voting outcome of
nonbinding shareholder proposals to adopt SOP but to alleviate these concerns they employ a
lsquofuzzyrsquo regression discontinuity design (RDD) essentially comparing the stock price reaction to
SOP proposals that pass by a small margin to the reaction to SOP proposals that fail by a small
10
Consistent with this interpretation the authors also find that SOP proposals generally targeted larger firms rather
than firms with excessive pay or lower pay-performance sensitivity However as noted by Ferri and Sandino (2009)
activists typically submit shareholder proposals aimed at promoting policy reforms at a broad sample of large firms
rather than the firms that would most benefit from the proposals because they believe that showing widespread
support for the proposal across all types of firms enhances its credibility with policy makers
margin (Cuntildeat Gine and Guadalupe 2012 Ertimur Ferri and Oesch forthcoming) The underlying idea
is that firms around the threshold are likely to have similar characteristics (indeed they do as the
authors show) but differ in the likelihood of implementation which is typically much higher for
proposals passing the threshold (Ertimur Ferri and Stubben 2010 Thomas and Cotter 2007)
Indeed the authors show that the likelihood of subsequently adopting SOP is 48 higher when
the SOP proposal passes by a small margin relatively to when it fails by a small margin Besides
because in these close-call situations the voting outcome is uncertain its resolution (pass or fail)
is likely to convey new information about the likelihood of SOP adoption making this setting
suitable to an event study11
Using this RDD approach Cuntildeat Gine and Guadalupe (2013) find that on the day of the vote a
SOP proposal that passes by a small margin yields an abnormal return of 24 relative to one
that fails (after controlling for other proposals voted upon at the same meeting) and estimate the
lsquofullrsquo value of SOP at 46 (after taking into account the increase in the probability of SOP
implementation)12
Aside from the issues discussed in Section IA (ie generalizability to other
firms) this estimate seems too large to reflect the present value of future reductions in excess
CEO pay (as noted by the authors their estimate of the value of SOP is equivalent to the
estimated value of removing two anti-takeover provision based on Cuntildeat Gine and Guadalupe
2012) particularly because (i) the authors find no evidence of subsequent changes in levels and
composition of CEO pay for firms implementing SOP (see Section IIA) an (ii) the sample firms
11
Consistent with this observation the authors find no market reaction to the voting outcome of SOP shareholder
proposals away from the threshold
12 Since the outcome of the vote is not binding the 24 market reaction only reflects the expected increase in the
probability of SOP adoption and thus understates the value of the SOP provision
do not appear to be characterized by excess CEO pay However the authors also find significant
improvements in subsequent operating performance and efficiency as a result of passing SOP
proposals and thus conjecture that the SOP vote is viewed as a tool to express a vote of
confidence in management performance more than a way to change compensation practices and
that the large positive market reaction reflects expected performance improvements under this
tighter monitoring regime While this is an interesting idea it remains unclear what additional
ldquoteethrdquo a SOP vote may provide over existing (non-compensation related) monitoring
mechanisms shareholders can express (and do often express) their dissatisfaction with
management performance when voting on director elections (Cai Garner and Walkling 2009)
C Event studies around compensation changes induced by SOP
Two studies examine announcements of changes to compensation plans related to SOP Larcker
McCall and Ormazabal (2013) examine the market reaction to (non-contaminated) compensation
changes disclosed in 8-K filings by Russell 3000 firms in the US during the year before the first
SOP vote They find that changes that appear to be made to avoid a negative proxy advisorrsquos
recommendation and thus a negative SOP vote are associated with a negative abnormal return
of -044 whereas other compensation changes are not associated with a significant stock price
reaction Ertimur Ferri and Oesch (2013) perform a similar analysis but focusing on
compensation changes made after the first SOP vote (and explicitly in response to the vote) by
firms receiving a negative recommendation and a high voting dissent in 2011 They fail to find
significant abnormal returns even for the subset of compensation changes that resulted in a
positive recommendation and low dissent in 2012 (and thus were presumably perceived to be
adequate and material by proxy advisors and voting shareholders)
D Other approaches effect of SOP on Tobinrsquos Q
In their cross-country study Correa and Lel (2013) also examine the effect of SOP laws on
Tobinrsquos Q and report a 36 increase in firm value following the adoption of the SOP laws
(more precisely a 36 higher firm value for post-SOP observations relative to a control sample
that pools together pre-SOP and non-SOP observations) They acknowledge that this increase in
firm value is too large to be justified by the relative decrease in CEO pay that they document and
suggest (but do not test) that it may reflect better alignment of pay and performance13
A key
concern is that higher valuation in countries with SOP laws may reflect other governance
changes introduced at the same time While the study controls for other compensation-related
laws there may be other non-compensation regulations introduced with SOP and with a
potentially larger impact on firm value
IV What have we learned
Adoption of SOP in the US has been long advocated by those who contend that executive pay is
a manifestation of rather than a solution to the agency problem and by those who favor great
shareholder involvement in corporate decisions A growing body of research is starting to
investigate the effect of SOP on executive pay and firm value It is premature to draw definitive
conclusions also because these studies differ in methodologies and settings but three points
seem to emerge First there is robust evidence that boards respond to SOP votes when
13
They also find that the firm value increase is higher when the relative decrease in CEO pay results in a lower pay
differential between CEO and other top managers (CEO pay slice) implying that a source of value creation may be
the reduced pay inequality among the top management team But this seems unlikely to explain the change in
Tobinrsquos Q since the increase occurs for both firms with advisory SOP laws and firms with binding SOP laws and
there is no change in CEO pay slice in firms subject to binding SOP laws
shareholders use it ldquovoicerdquo is heardrdquo Both in the US and UK studies show that not only firms
failing to win the SOP vote but also firms facing substantial dissent (20-30 of the votes
against) make changes to their compensation plans in consultation with institutional investors
and proxy advisors Second in both the US and UK institutional investors appear to use the
power of SOP votes to pressure firms into strengthening the link (or the perceived link) between
pay and performance (particularly on the downside) but not to pressure firms to reduce target
levels of pay suggesting a reluctance of institutional investors to lsquoregulatersquo executive pay Third
and consistent with the above most studies examining the aggregate effect of SOP on CEO pay
find some increase in pay-performance sensitivity but not effect on the level and growth of CEO
pay While some observers may interpret the lack of an effect on CEO pay levels as evidence of
the failure of SOP one should note that SOP per se is a neutral tool Its impact depends on what
investors choose to lsquosay on payrsquo Fourth based on my interpretation of the existing research to
date there is little evidence of an effect of SOP on firm value The studies documenting a
positive price impact appear subject to alternative interpretations or not plausible in their
findings (eg they fail to find an effect on CEO pay or the effect is too small to justify the
documented price impact hence it remains unclear what the source of value creation is) Also
there seems to be no stock price reaction to the SOP-induced compensation changes Finally
given the nature of these SOP-induced changes while some of them may be beneficial it seems
hard to believe that they will have a statistically detectable impact on firm value except perhaps
in a handful of firms where the compensation plan is subject to a complete overhaul
Perhaps one explanation for the weak impact of SOP is that executive pay problems were
overstated or that by the time SOP was introduced (more than a decade after the Enron-type
scandals that led to calls for greater shareholder voice) they had been already addressed via
other mechanisms (hedge fund activism monitoring by institutional investors vote-no
campaigns against compensation committee members SEC-mandated pay disclosures)
Overall though based on the evidence to date it does not appear that SOP had a major impact14
Was the adoption of SOP an lsquooptimalrsquo choice from a social welfare point of view This is a
difficult question to answer given the challenges of identifying and measuring all the potential
costs and benefits associated with SOP (likely to change over time and differ across countries)
Perhaps the most positive view of SOP is that its introduction prevented the adoption of other
more radical and intrusive regulatory measures (eg CEO pay caps) In that sense it may be
viewed as an lsquooptimalrsquo answer (ie the lesser of two evils) to the political pressure to reform
executive pay during the financial crisis
14
This interpretation of the evidence is also consistent with the prediction of analytical studies Ozbas and
Matsusaka (2013) model the benefits and costs of shareholder empowerment distinguishing between the right to
approve and the right to propose They show that permitting shareholders to propose directors or policies can cause
value-maximizing managers to take value-reducing actions to accommodate activist investors with non-value-
maximizing goals As for the right to approve it is weakly beneficial that is approval rights (such as a SOP vote)
are generally beneficial for shareholders in that they limit the managerrsquos ability to pursue private benefits at
shareholder expense but they have minimal impact on managerial actions and firm value (because the manager in
effect can threaten shareholders with an undesirable status quo if they do not approve the managerrsquos proposed
action)
References
Associated Press 2007 Administration opposes say on pay bill April 19
Bainbridge S 2008 Remarks on say on pay An unjustified incursion on director authority
UCLA School of Law Research Paper No 08-06
Bebchuk L 2005 The Case for Increasing Shareholder Power Harvard Law Review 118833-
917
Bebchuk L 2007 Written testimony submitted before the Committee on Financial Services
United States House of Representatives Hearing on Empowering Shareholders on Executive
Compensation March 8
Bebchuk LA A Cohen and A Ferrell 2009 What Matters in Corporate Governance Review
of Financial Studies 22783-827
Cai J J Garner and R Walkling 2009 Electing Directors Journal of Finance 642387-2419
Cai J and R Walkling 2011 Shareholdersrsquo Say on Pay Does it Create Value Journal of
Financial and Quantitative Analysis 46299ndash339
Coates IV JC (2009) Testimony before the Subcommittee on Securities Insurance and
Investment of the Committee on Banking Housing and Urban Affairs United States Senate
July 29 2009 httppapersssrncomsol3paperscfmabstract_id=1475355
Correa R and U Lel 2013 Say On Pay Laws Executive Compensation CEO Pay Slice and
Firm Value Around the World Federal Reserve Board Working Paper
Cuntildeat V M Gine and M Guadalupe 2012 The Vote is Cast The Effect of Corporate
Governance on Shareholder Value Journal of Finance 671943-1977
Cuntildeat V M Gine and M Guadalupe 2013 Say Pays Shareholder Voice and Firm
Performance ECGI - Finance Working Paper No 373
Del Guercio D L Seery and T Woidtke 2008 Do Boards Pay Attention When Institutional
Investors lsquoJust Vote Norsquo Journal of Financial Economics 9084-103
Ertimur Y F Ferri and S Stubben 2010 Board of Directorslsquo Responsiveness to Shareholders
Evidence from Shareholder Proposals Journal of Corporate Finance 1653-72
Ertimur Y F Ferri and V Muslu 2011 Shareholder Activism and CEO Pay Review of
Financial Studies 24535ndash592
Ertimur Y F Ferri and D Oesch 2013 Shareholder Votes and Proxy Advisors Evidence from
Say on Pay Journal of Accounting Research 51951-996
Ertimur Y F Ferri and D Oesch forthcoming Does the Director Election System Matter
Evidence from Majority Voting Review of Accounting Studies forthcoming
Ferri F 2012 Low-Cost Shareholder Activism A Review of the Evidence Research
Handbook on the Economics of Corporate Law ed CA Hill and BH McDonnell
Edward Elgar Publishing
Ferri F and D Maber 2013 Say on Pay Votes and CEO Compensation Evidence from the UK
Review of Finance 17527-563
Ferri F and D Oesch 2013 Management Influence on Investors Evidence from Shareholder
Votes on the Frequency of Say on Pay Columbia Business School Research Paper No 13-17
Ferri F and T Sandino 2009 The impact of shareholder activism on financial reporting and
compensation The case of employee stock options expensing The Accounting Review 84433ndash
466
Ferri F and J Weber 2009 AFSCME vs Mozilohellipand Say on Pay for All (A) Harvard
Business School Case 109-009
Gompers P J Ishii and A Metrick 2003 Corporate Governance and Equity Prices Quarterly
Journal of Economics 118107-155
Gordon J 2009 Say on Paylsquo Cautionary Notes on the UK Experience and the Case for
Shareholders Opt-In Harvard Journal on Legislation 46323-368
Iliev P and S Vitanova 2013 The effect of the Say on Pay in the US Pennsylvania State
University Working Paper
Kaplan S 2007 Testimony of Steven N Kaplan on Empowering Shareholders on Executive
Compensation and HR 1257 the Shareholder Vote on Executive Compensation Act before the
Committee on Financial Services United States House of Representatives March 8
Larcker D A McCall and G Ormazabal 2013 The Economic Consequences of Proxy
Advisor Say-on-Pay Voting Policies Working Paper Stanford University
Larcker D G Ormazabal and D Taylor 2011 The market reaction to corporate governance
regulation Journal of Financial Economics 101431ndash448
Levit D and N Malenko 2011 Non-binding voting for shareholder proposals Journal of
Finance 661579ndash1614
Mishel L and N Sabadish 2012 How executive compensation and financial-sector pay have
fueled income inequality Issue Brief 331 Economic Policy Institute
Murphy K 2012 The Politics of Pay A Legislative History of Executive Compensation in
Jennifer Hill and Randall Thomas eds Research Handbook on Executive Pay Edward Elgar
Publishers
New York Times 2007 House votes to give investors say on executive pay April 21
Norris F 2004 Corporate democracy and the power to embarrass New York Times March 4
Ozbas O and J Matsusaka2013 Managerial Accommodation Proxy Access and the Cost of
Shareholder Empowerment University of Southern California CLEO Research Paper Series No
C12-1
Thomas R and J Cotter 2007 Shareholder proposals in the new millennium Shareholder
support board response and market reaction Journal of Corporate Finance 13 368ndash391
Thomas R A Palmiter and J Cotter 2012 Dodd-Franks Say on Pay Will it Lead to a Greater
Role for Shareholders in Corporate Governance Cornell Law Review 971213-1266
Thomas R and C Van der Elst 2013 The International Scope of Say on Pay ECGI Law
Working Paper No227
compensation (hereafter SOP Bill)5 On the same day then-Senator Barack Obama introduced a
companion bill in the Senate (S1181)
But it was the financial crisis of 2007-2008 to accelerate the path toward mandatory
adoption of SOP Growing income inequality public outrage over Wall Street excesses (eg the
large bonuses paid at AIG) and banksrsquo bailouts and the perception that executive pay played a
role in inducing excessive risk-taking pressured policy-makers to take action with respect to
executive pay Many of the legislative proposals discussed in the House and Senate in 2008 and
2009 contained a SOP provision (see Table 1 in Larcker Ormazabal and Taylor 2011) Also
holding a SOP vote was a mandatory condition for firms to receive funds under the Troubled
Asset Relief Program (TARP) During the 2008 campaign both Presidential candidates
expressed support for SOP As these events unfolded the debate on merits and drawbacks of
SOP heated up Critics of SOP argued that at best SOP votes would be ignored (because
nonbinding) and at worst would cause directors to pander to shareholders with special interests
or lacking the required expertise and sophistication ultimately resulting in the adoption of
suboptimal pay practices (Kaplan 2007 Bainbridge 2008)6 They also cautioned that the high
5 House Bill 1257 Shareholder Vote on Executive Compensation Act
6 Analytical studies have also raised questions on the informativeness of SOP votes Ozbas and Matsusaka (2013)
note that SOP votes only reveal shareholder preferences for the current compensation plan compared to the
alternative plan that will be adopted if the current plan is rejected but such alternative plan is unclear in practice
casting doubts on whether the shareholder votes on SOP may be truly informative Levit and Malenko (2011) further
suggest that shareholder votes may not be informative about shareholder preferences when shareholders ignore their
own information and vote strategically conditioning on the chance of casting a pivotal vote
cost of analyzing executive pay at thousands of firms would lead shareholders to outsource
voting decisions to proxy advisors who in turn would minimize their own costs by promoting
one-size-fits-all compensation practices which would hurt firm value Finally SOP opponents
noted that shareholders had already tools to express their views on compensation matters such as
the ability to submit shareholder proposals on compensation and withhold votes from directors
responsible for compensation packages Supporters of SOP argued that enhanced shareholder
voice (as formalized in a SOP vote) and reputation concerns would help boards overcome
psychological barriers to negotiating with CEOs on behalf of shareholders resulting in more
efficient compensation contracts and more dialogue between boards and shareholders (Bebchuk
2007) They also pointed out to growing evidence of shareholdersrsquo sophistication in casting
informed votes Finally they argued that SOP would be more effective than shareholder
proposals (limited to a single issue) and less disruptive than a confrontational vote against an
otherwise valuable director More fundamentally critics and supporters of SOP disagreed on
whether existing compensation contracts were the result of an efficient labor market for talent or
instead the expression of management power over captive boards
Finally in 2010 a SOP provision found its way in the Dodd-Frank Act Ironically SOP a
provision designed to provide more alignment between shareholder and management was
adopted as part of a financial reform package aimed at deterring lsquoexcessiversquo risk taking
(excessive from the point of view of social welfare) which may be the result of the alignment of
interests between shareholders and management
As of today non-binding SOP votes (on the compensation report) have been mandated in
Australia Belgium Denmark Israel Italy Portugal while Netherlands Sweden Norway South
Africa and Switzerland adopted some version of a binding SOP vote (on the future compensation
policy) which is now also being considered by the European Union7
When putting SOP into its historical context it is also important to appreciate that SOP has come
to epitomize a broader movement toward greater shareholder democracy beyond the issue of
executive pay For example during the same period shareholder activists have long
(unsuccessfully) lobbied for a proxy access rule that would have made it easier for shareholders
to oust corporate directors and nominate their own candidates Hence many observers often
view the successes and failures of SOP as speaking to the potential effects of other reforms
aimed at increasing shareholder power making the academic research on SOP all more relevant
II The Effect of Say on Pay on Executive Compensation
A Effect of SOP votes on level and composition of executive pay
As discussed earlier United Kingdom was the first country to adopt SOP in 2002 for publicly
traded firms Ferri and Maber (2013) report that in most cases shareholders voted in favor of
compensation plans Failed SOP votes (ie greater than 50 of votes against) were rare (2 of
the sample) though highly publicized in the press However about one-fourth of the sample
firms received more than 20 of votes against
To capture the overall effect of SOP on CEO pay Ferri and Maber (2013) examine the
sensitivity of CEO pay to its economic determinants over the 2000-2005 period (ie before and
after the introduction of SOP regulation) and document a significant increase in the sensitivity of
7 Some countries have already modified their first SOP legislation In 2013 the UK added to the non-binding
backward-looking SOP vote on the current compensation report a binding forward-looking vote on the proposed
compensation policy In 2012 Australia modified the non-binding SOP vote to introduce the so-called ldquotwo-strikesrdquo
rule Under the rule if 25 of shareholders vote against a companyrsquos remuneration report at two consecutive annual
general meetings the entire board may have to stand for re-election within three months For more details about the
international development of SOP see Thomas and Van der Elst (2013)
CEO pay to poor performance They also find that this increase does not occur for a subset of
UK firms exempted from the SOP regulation (firms traded on the Alternative Investment Market
a sub-market of the London Stock Exchange with a more flexible regulatory system) suggesting
that the result reflects the impact of SOP rather than a general trend affecting all firms Besides
Ferri and Maber (2013) find that the increase is more pronounced in firms experiencing high
voting dissent and firms with high abnormal CEO pay before the adoption of SOP consistent
with a causal impact of SOP In contrast they fail to find any effect of SOP on the level of pay
Correa and Lel (2013) examine the effect of SOP laws on CEO pay using a large cross-country
sample of about 103000 firm-year observations from 39 countries including 12 countries that
adopted some (advisory or binding) version of SOP They essentially compare post-SOP CEO
pay at firms of countries adopting SOP to a control sample which includes all non-SOP
observations (that is pre-SOP CEO pay in countries eventually adopting SOP as well as CEO
pay in countries never adopting SOP) They find that firms in a post-SOP regime (i) exhibit
lower CEO pay levels (but only in countries adopting an advisory SOP vote) a finding driven by
a relative decline in equity awards and limited to CEOs and not the other top executives (ii)
higher pay-performance sensitivity (the authors do not look at positive and negative performance
separately) The current version of the paper does not examine whether these findings are more
pronounced in (or driven by) firms with excess CEO pay before the adoption of SOP and in
firms experiencing adverse SOP votes
The drawback of the research design in Correa and Lel (2013) is that it is not clear whether the
results reflect time-series changes within countries adopting SOP (pre vs post) or difference
between firms in SOP and non-SOP countries For example further tests in the paper suggest
that the lower CEO pay levels for post-SOP observations are not driven by a pay decrease
relative to the pre-SOP period (in fact country-by-country analyses indicate no change in CEO
pay levels after the adoption of SOP) but by the comparison with non-SOP countries Also while
the authors employ reasonable econometric solutions to deal with this issue some concerns
remains as to whether non-SOP countries can be a good benchmark for what would have
happened in SOP countries had SOP not been adopted (in other words if only countries with
problematic CEO practices adopt SOP it is possible that changes in CEO pay reflect factors
leading to the adoption of SOP rather than the effect of SOP laws per se) Notwithstanding these
concerns these findings are intriguing and future versions of this recent study have the potential
to provide more conclusive evidence on the effect of SOP
While in the US SOP was mandated by the Dodd-Frank Act its implementation was left to the
Securities and Exchange Commission (SEC) The final rule issued in January 2011 exempted
lsquosmallrsquo firms (ie firms with a public float below $75 million) from the SOP provision for two
years Iliev and Vitanova (2013) exploit this setting to estimate the effect of SOP on CEO
compensation In particularly using a regression discontinuity design they compare changes in
CEO pay of firms just above the SEC-imposed threshold to changes in CEO pay of firms just
below the threshold (but otherwise quite similar) They find no differences in terms of level and
composition of CEO pay and in terms of golden parachutes concluding that SOP had no impact
on CEO pay While the identification strategy chosen by the authors has the benefit of reducing
endogeneity issues it comes with the price of focusing only on fairly small firms (those around
the $75 million public float threshold) where CEO pay may not be a problem in the first place
(in which case the lack of impact of SOP would not be surprising but would not speak to its
impact at large firms traditionally the target of compensation-related criticism)
Cuntildeat Gine and Guadalupe (2013) also use a regression discontinuity design but in a different
setting In particular they analyze nonbinding shareholder proposals to adopt SOP (250
proposals between 2006 and 2010) Comparing the changes in CEO pay for firms voluntarily
adopting SOP in response to the proposals to firms not adopting SOP would be difficult since
the adoption decision is endogenous Similarly it would be difficult to compare changes in CEO
pay between firms where the proposal is passed (and thus significantly more likely to be
adopted) and firms where the proposal fails to be approved since the voting outcome is
endogenous as well
Hence Cuntildeat Gine and Guadalupe (2013) use instead a regression discontinuity design that
compares firms where the proposals receive slightly more than 50 of the votes (the threshold
for approval) to otherwise similar firms where the proposals receive slightly less than 50
Using this approach they conclude that the passing and adoption of SOP proposals is not
associated with changes in the level of CEO pay nor its composition and structure (mix of cash
and equity equity incentives level of perks and deferred pay) In some sense this finding
complements the one in Iliev and Vitanova (2013) in that it is based on large firms (shareholder
proposals are typically submitted at large SampP 500 firms) However both Cai and Walkling
(2011) and Cuntildeat Gine and Guadalupe (2013) show that firms targeted by SOP proposals do not
exhibit excess CEO pay or weaker governance8 Hence the lack of an effect at these firms may
8 Activists submitting SOP proposals chose to focus on a broad sample of large firms rather than target only firms
with problematic CEO pay practices to obtain visibility and show to policy-makers that investorrsquos support for SOP
was not limited to problematic firms (Ferri and Weber 2009)
not be surprising More generally the findings cannot be necessarily generalized to firms away
from the 50 threshold or to firms not targeted by SOP shareholder proposals
B Effect of SOP votes on compensation practices
Analyses of the effect of SOP on level and composition of CEO pay may not capture a number
of other changes to compensation contracts that may be induced by SOP votes but are not
reflected in measure of CEO pay For example the introduction of performance-based vesting
provisions is typically not reflected in the estimates of fair value of equity grants used by
researchers in measuring CEO pay Many other changes (eg terms of severance packages) will
only be reflected in measures of CEO pay contingent upon the occurrence of certain events
Hence it is important to complement the evidence in section IIA with an analysis of firmsrsquo
disclosures of changes to observable provisions of compensation contracts Another benefit of
this approach is that it captures the changes that boards explicitly present as a response to SOP
votes and thus it acts as a reality check on regression-based inferences For example a
regression-based finding of say a decrease in CEO pay level would be more credible if
supported by evidence that firms disclose a decision to reduce target levels of CEO pay in
response to an adverse SOP vote (an action that presumably they have an incentive to disclose to
get rewarded by shareholders in subsequent votes)
Two studies collect data on specific changes made to compensation contracts explicitly in
response to SOP votes in the UK and the US using firmsrsquo disclosures in their proxy statements
Ferri and Maber (2013) examine the compensation reports of a sample of UK firms before and
after the first SOP vote and find that firms experiencing higher voting dissent were significantly
more likely to remove compensation provisions criticized by investors as lsquorewards for failurersquo
relative to a matched sample of firms experiencing lower dissent For example they report that
among firms with long notice periods (implying larger severance payments ) the percentage of
high dissent firms that shortened them after the vote (800) was significantly higher than before
the vote (200) and also significantly higher than among low dissent firms after the vote
(333) They report similar findings for another controversial practice namely the presence of
retesting provisions in the performance-based vesting conditions of equity grants Most firms
indicate that these changes were the result of consultations with their major institutional
investors Firmsrsquo responsiveness to SOP votes was lsquorewardedrsquo with a substantial increase in
favorable SOP votes at the subsequent annual meeting Ferri and Maber (2013) also find that
many firms experiencing low voting dissent at the first SOP vote had removed these practices
before the vote9 highlighting the importance of accounting for ex ante effects when assessing the
impact of a new regulation
Ertimur Ferri and Oesch (2013) examine the effect of SOP on compensation practices in the US
in 2011 (first proxy season under mandatory SOP) Their key findings are generally similar to
the evidence from the UK (i) failed SOP votes are rare (about 2 of the sample) though cases
of substantial dissent are more frequent (ii) more than 55 of the firms experiencing significant
voting dissent respond by making material changes to their compensation plan during the
subsequent year (eg introduction of performance-based vesting conditions in equity grants use
of tougher performance targets removal of perks and tax gross-ups removal of controversial
provisions from severance contracts etc) usually in consultation with major institutional
9 In their sample 700 of low dissent firms shortened their notice periods during the year before the SOP vote
investors (iii) firms making changes to their compensation plans receive greater voting support
at the following SOP vote
Ertimur Ferri and Oesch (2013) also highlight the strong influence of the recommendations
released by proxy advisors (particularly Institutional Shareholder Services ISS) with a negative
recommendation being associated with 25 more votes against the compensation plan Further
evidence of the significant influence of ISS is the striking discontinuity in the relation between
the extent of SOP voting dissent and firmsrsquo responsiveness to the vote the percentage of firms
making compensation changes in response to SOP votes jumps from 32 to 72 around the
30 voting dissent threshold Why After the 2011 proxy season ISS had indicated that firms
failing to ldquoadequatelyrdquo respond to SOP voting dissent above 30 would receive a negative
recommendation in 2012 on the SOP proposal and on the election of compensation committee
members
Finally similar to the UK experience there is also anecdotal evidence of analogous
compensation changes being made by US firms ahead of the SOP vote to avoid a negative
proxy advisor recommendation and an adverse shareholder vote (Thomas Palmiter and Cotter
2013 Larcker McCall and Ormazabal 2013)
III The Effect of Say on Pay on Firm Value
In this section we examine the empirical studies on the overall effect of SOP on firm value
A Event studies around the adoption of Say on Pay regulations
As mentioned in the Introduction on April 20 2007 the House of Representatives passed a SOP
Bill by a 2-1 margin On the same day then-Senator Barack Obama introduced a companion
bill (S1181) in the Senate (which was then put on hold by the Senate Banking Committee)
While the SOP Billrsquos approval was expected (Democrats were in control of the House and
supported the SOP Bill) the 2-1 margin was unexpected suggesting some support for SOP
among Republicans
Cai and Walkling (2011) examine the market reaction to the Housersquos approval of the SOP Bill
for a sample of firms in the SampP 1500 index and document a positive reaction for firms more
likely to benefit from greater shareholder voice over executive pay namely firms with high
abnormal CEO cash pay (defined as the difference between actual CEO cash pay and the level
predicted based on known economic determinants such a size performance industry) firms
with low pay-for-performance sensitivity firms with a history of shareholders willing to vote
against compensation-related management proposals and firms with a history of responsiveness
to shareholder pressure on compensation issues In other words they find a positive reaction in
firms where the compensation problems are more severe and a say on pay vote is more likely to
trigger a response (eg more shareholders willing to vote against management and greater firmsrsquo
propensity to respond to an adverse vote) Note however that they do not find a significant
impact for firms with high abnormal equity and total CEO pay
Larcker Ormazabal and Taylor (2011) also examine the market reaction to the Housersquos approval
of the SOP Bill (in addition to other regulatory events related to executive pay and other
governance provisions) Similar to Cai and Wakling (2011) they fail to find a significant impact
for firms with high abnormal total CEO pay (they do not examine the price reaction for firms
with abnormal cash CEO pay or firms with lower pay-performance sensitivity)
A concern with both studies is that the Housersquos approval of a bill does not guarantee passage in
the Senate and approval by the White House In fact the press at the time reported that the
prospects of the bill in the Senate were uncertain (New York Times 2007) and the Bush White
House openly opposed the Bill (Associated Press 2007) Hence it is not clear the extent to
which this event increased the likelihood of a say on pay legislation Arguably as noted in
Section I the only lsquoeventrsquo that substantially increased the likelihood of say on pay legislation
was the financial crisis and the related perception of compensation abuses (eg outrage over AIG
bonuses)
Ferri and Maber (2013) argue that the announcement of the submission of SOP regulation to the
Parliament in the United Kingdom in June 2002 offers a more powerful setting for an event
study since it was largely unexpected and increased substantially the probability of SOP
adoption (Parliamentrsquos approval was virtually guaranteed) Using this event they document
positive abnormal returns for firms with excess CEO pay combined with poor performance and
for firms with controversial pay practices that weaken the penalties for poor performance (those
practices were often removed in response to SOP votes as discussed earlier in Section IIB)
They interpret their findings as consistent with shareholders perceiving SOP as a value
enhancing monitoring mechanism for firms with low pay-to-poor-performance sensitivity
Finally a recent study by Iliev and Vitanova (2013) examines the market reaction around the
announcement of the SEC final SOP rule that on January 25 2011 exempted small firms (ie
firms with a public float below $75 million) from adopting SOP for two years (press reports at
the time noted that the exemption could become permanent) An appealing feature of this setting
is that the SEC decision was an unexpected reversal from the rule proposed in October 2010 rule
under which no publicly traded firm would be exempted
Iliev and Vitanova (2013) compare the market reaction to this announcement for firms just above
the SEC-imposed threshold and thus subject to the SOP rule and firms just below the threshold
(but otherwise quite similar) and thus exempted from SOP While they find not significant
abnormal returns for either group the abnormal returns for exempted firms are significantly
more negative a finding that the authors interpret as evidence of a positive value effect of
mandatory SOP While the setting is an interesting one some concerns remain as exempted
firms are fairly small and the magnitude of the executive pay problem (if any) is unlikely to be
large enough to explain the differential market reaction Also the same study finds no effect on
CEO pay level and composition (see Section IIA) Hence the reason for the differential stock
price reaction remains unclear It would be helpful to know if the event study result is driven by
(or stronger for) exempted firms with excessive pay or problematic pay practices
B Event studies around Shareholder Proposals to Adopt Say on Pay
Starting in 2006 shareholder activists led by union pension funds submitted shareholder
proposal to adopt SOP at hundreds of US firms in an attempt to induce voluntary or mandatory
widespread adoption of SOP
Cai and Walkling (2011) examine the market reaction to proxy filings and annual meetings of
113 firms targeted by nonbinding shareholder proposals to adopt SOP between 2006 and 2008
On average they find insignificant returns around both the submission of the proposal (proxy
filing date) and the vote on the proposal (annual meeting date) But they also find that when the
proposals are filed by union pension funds the abnormal returns (still insignificant) are more
negative than when filed by other activists and that when the proposal is defeated the abnormal
returns (while insignificant) are more positive than when the proposal is passed (it is not clear
whether this result is driven by union-sponsored proposals) They interpret these findings as
evidence that the market views SOP proposals filed by union pension funds are driven by special
interests rather than value maximization10
However caution is required in interpreting this
evidence Virtually all activists filing SOP proposals were coordinated by a group of investors
(mostly but not only union pension funds) who made available to other activists a template for
SOP proposals and a list of target firms (Ferri and Weber 2009) Hence the distinction between
union and non-union proponents in the context of SOP proposals may be only apparent More
importantly the list of target firms was publicly available months before the proxy filing dates
so it is not clear whether the proxy statements contained any new information Similarly the
voting outcome of the SOP proposals may largely be anticipated (based on the composition of
institutional owners proxy advisorsrsquo recommendations etc) Finally the analyses do not control
for other (potentially new) information contained in proxy statements (eg executive
compensation report other items up for a vote at the annual meeting) or other events occurring at
the annual meeting (eg shareholder votes on other items)
Cuntildeat Gine and Guadalupe (2013) also examine the market reaction to the voting outcome of
nonbinding shareholder proposals to adopt SOP but to alleviate these concerns they employ a
lsquofuzzyrsquo regression discontinuity design (RDD) essentially comparing the stock price reaction to
SOP proposals that pass by a small margin to the reaction to SOP proposals that fail by a small
10
Consistent with this interpretation the authors also find that SOP proposals generally targeted larger firms rather
than firms with excessive pay or lower pay-performance sensitivity However as noted by Ferri and Sandino (2009)
activists typically submit shareholder proposals aimed at promoting policy reforms at a broad sample of large firms
rather than the firms that would most benefit from the proposals because they believe that showing widespread
support for the proposal across all types of firms enhances its credibility with policy makers
margin (Cuntildeat Gine and Guadalupe 2012 Ertimur Ferri and Oesch forthcoming) The underlying idea
is that firms around the threshold are likely to have similar characteristics (indeed they do as the
authors show) but differ in the likelihood of implementation which is typically much higher for
proposals passing the threshold (Ertimur Ferri and Stubben 2010 Thomas and Cotter 2007)
Indeed the authors show that the likelihood of subsequently adopting SOP is 48 higher when
the SOP proposal passes by a small margin relatively to when it fails by a small margin Besides
because in these close-call situations the voting outcome is uncertain its resolution (pass or fail)
is likely to convey new information about the likelihood of SOP adoption making this setting
suitable to an event study11
Using this RDD approach Cuntildeat Gine and Guadalupe (2013) find that on the day of the vote a
SOP proposal that passes by a small margin yields an abnormal return of 24 relative to one
that fails (after controlling for other proposals voted upon at the same meeting) and estimate the
lsquofullrsquo value of SOP at 46 (after taking into account the increase in the probability of SOP
implementation)12
Aside from the issues discussed in Section IA (ie generalizability to other
firms) this estimate seems too large to reflect the present value of future reductions in excess
CEO pay (as noted by the authors their estimate of the value of SOP is equivalent to the
estimated value of removing two anti-takeover provision based on Cuntildeat Gine and Guadalupe
2012) particularly because (i) the authors find no evidence of subsequent changes in levels and
composition of CEO pay for firms implementing SOP (see Section IIA) an (ii) the sample firms
11
Consistent with this observation the authors find no market reaction to the voting outcome of SOP shareholder
proposals away from the threshold
12 Since the outcome of the vote is not binding the 24 market reaction only reflects the expected increase in the
probability of SOP adoption and thus understates the value of the SOP provision
do not appear to be characterized by excess CEO pay However the authors also find significant
improvements in subsequent operating performance and efficiency as a result of passing SOP
proposals and thus conjecture that the SOP vote is viewed as a tool to express a vote of
confidence in management performance more than a way to change compensation practices and
that the large positive market reaction reflects expected performance improvements under this
tighter monitoring regime While this is an interesting idea it remains unclear what additional
ldquoteethrdquo a SOP vote may provide over existing (non-compensation related) monitoring
mechanisms shareholders can express (and do often express) their dissatisfaction with
management performance when voting on director elections (Cai Garner and Walkling 2009)
C Event studies around compensation changes induced by SOP
Two studies examine announcements of changes to compensation plans related to SOP Larcker
McCall and Ormazabal (2013) examine the market reaction to (non-contaminated) compensation
changes disclosed in 8-K filings by Russell 3000 firms in the US during the year before the first
SOP vote They find that changes that appear to be made to avoid a negative proxy advisorrsquos
recommendation and thus a negative SOP vote are associated with a negative abnormal return
of -044 whereas other compensation changes are not associated with a significant stock price
reaction Ertimur Ferri and Oesch (2013) perform a similar analysis but focusing on
compensation changes made after the first SOP vote (and explicitly in response to the vote) by
firms receiving a negative recommendation and a high voting dissent in 2011 They fail to find
significant abnormal returns even for the subset of compensation changes that resulted in a
positive recommendation and low dissent in 2012 (and thus were presumably perceived to be
adequate and material by proxy advisors and voting shareholders)
D Other approaches effect of SOP on Tobinrsquos Q
In their cross-country study Correa and Lel (2013) also examine the effect of SOP laws on
Tobinrsquos Q and report a 36 increase in firm value following the adoption of the SOP laws
(more precisely a 36 higher firm value for post-SOP observations relative to a control sample
that pools together pre-SOP and non-SOP observations) They acknowledge that this increase in
firm value is too large to be justified by the relative decrease in CEO pay that they document and
suggest (but do not test) that it may reflect better alignment of pay and performance13
A key
concern is that higher valuation in countries with SOP laws may reflect other governance
changes introduced at the same time While the study controls for other compensation-related
laws there may be other non-compensation regulations introduced with SOP and with a
potentially larger impact on firm value
IV What have we learned
Adoption of SOP in the US has been long advocated by those who contend that executive pay is
a manifestation of rather than a solution to the agency problem and by those who favor great
shareholder involvement in corporate decisions A growing body of research is starting to
investigate the effect of SOP on executive pay and firm value It is premature to draw definitive
conclusions also because these studies differ in methodologies and settings but three points
seem to emerge First there is robust evidence that boards respond to SOP votes when
13
They also find that the firm value increase is higher when the relative decrease in CEO pay results in a lower pay
differential between CEO and other top managers (CEO pay slice) implying that a source of value creation may be
the reduced pay inequality among the top management team But this seems unlikely to explain the change in
Tobinrsquos Q since the increase occurs for both firms with advisory SOP laws and firms with binding SOP laws and
there is no change in CEO pay slice in firms subject to binding SOP laws
shareholders use it ldquovoicerdquo is heardrdquo Both in the US and UK studies show that not only firms
failing to win the SOP vote but also firms facing substantial dissent (20-30 of the votes
against) make changes to their compensation plans in consultation with institutional investors
and proxy advisors Second in both the US and UK institutional investors appear to use the
power of SOP votes to pressure firms into strengthening the link (or the perceived link) between
pay and performance (particularly on the downside) but not to pressure firms to reduce target
levels of pay suggesting a reluctance of institutional investors to lsquoregulatersquo executive pay Third
and consistent with the above most studies examining the aggregate effect of SOP on CEO pay
find some increase in pay-performance sensitivity but not effect on the level and growth of CEO
pay While some observers may interpret the lack of an effect on CEO pay levels as evidence of
the failure of SOP one should note that SOP per se is a neutral tool Its impact depends on what
investors choose to lsquosay on payrsquo Fourth based on my interpretation of the existing research to
date there is little evidence of an effect of SOP on firm value The studies documenting a
positive price impact appear subject to alternative interpretations or not plausible in their
findings (eg they fail to find an effect on CEO pay or the effect is too small to justify the
documented price impact hence it remains unclear what the source of value creation is) Also
there seems to be no stock price reaction to the SOP-induced compensation changes Finally
given the nature of these SOP-induced changes while some of them may be beneficial it seems
hard to believe that they will have a statistically detectable impact on firm value except perhaps
in a handful of firms where the compensation plan is subject to a complete overhaul
Perhaps one explanation for the weak impact of SOP is that executive pay problems were
overstated or that by the time SOP was introduced (more than a decade after the Enron-type
scandals that led to calls for greater shareholder voice) they had been already addressed via
other mechanisms (hedge fund activism monitoring by institutional investors vote-no
campaigns against compensation committee members SEC-mandated pay disclosures)
Overall though based on the evidence to date it does not appear that SOP had a major impact14
Was the adoption of SOP an lsquooptimalrsquo choice from a social welfare point of view This is a
difficult question to answer given the challenges of identifying and measuring all the potential
costs and benefits associated with SOP (likely to change over time and differ across countries)
Perhaps the most positive view of SOP is that its introduction prevented the adoption of other
more radical and intrusive regulatory measures (eg CEO pay caps) In that sense it may be
viewed as an lsquooptimalrsquo answer (ie the lesser of two evils) to the political pressure to reform
executive pay during the financial crisis
14
This interpretation of the evidence is also consistent with the prediction of analytical studies Ozbas and
Matsusaka (2013) model the benefits and costs of shareholder empowerment distinguishing between the right to
approve and the right to propose They show that permitting shareholders to propose directors or policies can cause
value-maximizing managers to take value-reducing actions to accommodate activist investors with non-value-
maximizing goals As for the right to approve it is weakly beneficial that is approval rights (such as a SOP vote)
are generally beneficial for shareholders in that they limit the managerrsquos ability to pursue private benefits at
shareholder expense but they have minimal impact on managerial actions and firm value (because the manager in
effect can threaten shareholders with an undesirable status quo if they do not approve the managerrsquos proposed
action)
References
Associated Press 2007 Administration opposes say on pay bill April 19
Bainbridge S 2008 Remarks on say on pay An unjustified incursion on director authority
UCLA School of Law Research Paper No 08-06
Bebchuk L 2005 The Case for Increasing Shareholder Power Harvard Law Review 118833-
917
Bebchuk L 2007 Written testimony submitted before the Committee on Financial Services
United States House of Representatives Hearing on Empowering Shareholders on Executive
Compensation March 8
Bebchuk LA A Cohen and A Ferrell 2009 What Matters in Corporate Governance Review
of Financial Studies 22783-827
Cai J J Garner and R Walkling 2009 Electing Directors Journal of Finance 642387-2419
Cai J and R Walkling 2011 Shareholdersrsquo Say on Pay Does it Create Value Journal of
Financial and Quantitative Analysis 46299ndash339
Coates IV JC (2009) Testimony before the Subcommittee on Securities Insurance and
Investment of the Committee on Banking Housing and Urban Affairs United States Senate
July 29 2009 httppapersssrncomsol3paperscfmabstract_id=1475355
Correa R and U Lel 2013 Say On Pay Laws Executive Compensation CEO Pay Slice and
Firm Value Around the World Federal Reserve Board Working Paper
Cuntildeat V M Gine and M Guadalupe 2012 The Vote is Cast The Effect of Corporate
Governance on Shareholder Value Journal of Finance 671943-1977
Cuntildeat V M Gine and M Guadalupe 2013 Say Pays Shareholder Voice and Firm
Performance ECGI - Finance Working Paper No 373
Del Guercio D L Seery and T Woidtke 2008 Do Boards Pay Attention When Institutional
Investors lsquoJust Vote Norsquo Journal of Financial Economics 9084-103
Ertimur Y F Ferri and S Stubben 2010 Board of Directorslsquo Responsiveness to Shareholders
Evidence from Shareholder Proposals Journal of Corporate Finance 1653-72
Ertimur Y F Ferri and V Muslu 2011 Shareholder Activism and CEO Pay Review of
Financial Studies 24535ndash592
Ertimur Y F Ferri and D Oesch 2013 Shareholder Votes and Proxy Advisors Evidence from
Say on Pay Journal of Accounting Research 51951-996
Ertimur Y F Ferri and D Oesch forthcoming Does the Director Election System Matter
Evidence from Majority Voting Review of Accounting Studies forthcoming
Ferri F 2012 Low-Cost Shareholder Activism A Review of the Evidence Research
Handbook on the Economics of Corporate Law ed CA Hill and BH McDonnell
Edward Elgar Publishing
Ferri F and D Maber 2013 Say on Pay Votes and CEO Compensation Evidence from the UK
Review of Finance 17527-563
Ferri F and D Oesch 2013 Management Influence on Investors Evidence from Shareholder
Votes on the Frequency of Say on Pay Columbia Business School Research Paper No 13-17
Ferri F and T Sandino 2009 The impact of shareholder activism on financial reporting and
compensation The case of employee stock options expensing The Accounting Review 84433ndash
466
Ferri F and J Weber 2009 AFSCME vs Mozilohellipand Say on Pay for All (A) Harvard
Business School Case 109-009
Gompers P J Ishii and A Metrick 2003 Corporate Governance and Equity Prices Quarterly
Journal of Economics 118107-155
Gordon J 2009 Say on Paylsquo Cautionary Notes on the UK Experience and the Case for
Shareholders Opt-In Harvard Journal on Legislation 46323-368
Iliev P and S Vitanova 2013 The effect of the Say on Pay in the US Pennsylvania State
University Working Paper
Kaplan S 2007 Testimony of Steven N Kaplan on Empowering Shareholders on Executive
Compensation and HR 1257 the Shareholder Vote on Executive Compensation Act before the
Committee on Financial Services United States House of Representatives March 8
Larcker D A McCall and G Ormazabal 2013 The Economic Consequences of Proxy
Advisor Say-on-Pay Voting Policies Working Paper Stanford University
Larcker D G Ormazabal and D Taylor 2011 The market reaction to corporate governance
regulation Journal of Financial Economics 101431ndash448
Levit D and N Malenko 2011 Non-binding voting for shareholder proposals Journal of
Finance 661579ndash1614
Mishel L and N Sabadish 2012 How executive compensation and financial-sector pay have
fueled income inequality Issue Brief 331 Economic Policy Institute
Murphy K 2012 The Politics of Pay A Legislative History of Executive Compensation in
Jennifer Hill and Randall Thomas eds Research Handbook on Executive Pay Edward Elgar
Publishers
New York Times 2007 House votes to give investors say on executive pay April 21
Norris F 2004 Corporate democracy and the power to embarrass New York Times March 4
Ozbas O and J Matsusaka2013 Managerial Accommodation Proxy Access and the Cost of
Shareholder Empowerment University of Southern California CLEO Research Paper Series No
C12-1
Thomas R and J Cotter 2007 Shareholder proposals in the new millennium Shareholder
support board response and market reaction Journal of Corporate Finance 13 368ndash391
Thomas R A Palmiter and J Cotter 2012 Dodd-Franks Say on Pay Will it Lead to a Greater
Role for Shareholders in Corporate Governance Cornell Law Review 971213-1266
Thomas R and C Van der Elst 2013 The International Scope of Say on Pay ECGI Law
Working Paper No227
cost of analyzing executive pay at thousands of firms would lead shareholders to outsource
voting decisions to proxy advisors who in turn would minimize their own costs by promoting
one-size-fits-all compensation practices which would hurt firm value Finally SOP opponents
noted that shareholders had already tools to express their views on compensation matters such as
the ability to submit shareholder proposals on compensation and withhold votes from directors
responsible for compensation packages Supporters of SOP argued that enhanced shareholder
voice (as formalized in a SOP vote) and reputation concerns would help boards overcome
psychological barriers to negotiating with CEOs on behalf of shareholders resulting in more
efficient compensation contracts and more dialogue between boards and shareholders (Bebchuk
2007) They also pointed out to growing evidence of shareholdersrsquo sophistication in casting
informed votes Finally they argued that SOP would be more effective than shareholder
proposals (limited to a single issue) and less disruptive than a confrontational vote against an
otherwise valuable director More fundamentally critics and supporters of SOP disagreed on
whether existing compensation contracts were the result of an efficient labor market for talent or
instead the expression of management power over captive boards
Finally in 2010 a SOP provision found its way in the Dodd-Frank Act Ironically SOP a
provision designed to provide more alignment between shareholder and management was
adopted as part of a financial reform package aimed at deterring lsquoexcessiversquo risk taking
(excessive from the point of view of social welfare) which may be the result of the alignment of
interests between shareholders and management
As of today non-binding SOP votes (on the compensation report) have been mandated in
Australia Belgium Denmark Israel Italy Portugal while Netherlands Sweden Norway South
Africa and Switzerland adopted some version of a binding SOP vote (on the future compensation
policy) which is now also being considered by the European Union7
When putting SOP into its historical context it is also important to appreciate that SOP has come
to epitomize a broader movement toward greater shareholder democracy beyond the issue of
executive pay For example during the same period shareholder activists have long
(unsuccessfully) lobbied for a proxy access rule that would have made it easier for shareholders
to oust corporate directors and nominate their own candidates Hence many observers often
view the successes and failures of SOP as speaking to the potential effects of other reforms
aimed at increasing shareholder power making the academic research on SOP all more relevant
II The Effect of Say on Pay on Executive Compensation
A Effect of SOP votes on level and composition of executive pay
As discussed earlier United Kingdom was the first country to adopt SOP in 2002 for publicly
traded firms Ferri and Maber (2013) report that in most cases shareholders voted in favor of
compensation plans Failed SOP votes (ie greater than 50 of votes against) were rare (2 of
the sample) though highly publicized in the press However about one-fourth of the sample
firms received more than 20 of votes against
To capture the overall effect of SOP on CEO pay Ferri and Maber (2013) examine the
sensitivity of CEO pay to its economic determinants over the 2000-2005 period (ie before and
after the introduction of SOP regulation) and document a significant increase in the sensitivity of
7 Some countries have already modified their first SOP legislation In 2013 the UK added to the non-binding
backward-looking SOP vote on the current compensation report a binding forward-looking vote on the proposed
compensation policy In 2012 Australia modified the non-binding SOP vote to introduce the so-called ldquotwo-strikesrdquo
rule Under the rule if 25 of shareholders vote against a companyrsquos remuneration report at two consecutive annual
general meetings the entire board may have to stand for re-election within three months For more details about the
international development of SOP see Thomas and Van der Elst (2013)
CEO pay to poor performance They also find that this increase does not occur for a subset of
UK firms exempted from the SOP regulation (firms traded on the Alternative Investment Market
a sub-market of the London Stock Exchange with a more flexible regulatory system) suggesting
that the result reflects the impact of SOP rather than a general trend affecting all firms Besides
Ferri and Maber (2013) find that the increase is more pronounced in firms experiencing high
voting dissent and firms with high abnormal CEO pay before the adoption of SOP consistent
with a causal impact of SOP In contrast they fail to find any effect of SOP on the level of pay
Correa and Lel (2013) examine the effect of SOP laws on CEO pay using a large cross-country
sample of about 103000 firm-year observations from 39 countries including 12 countries that
adopted some (advisory or binding) version of SOP They essentially compare post-SOP CEO
pay at firms of countries adopting SOP to a control sample which includes all non-SOP
observations (that is pre-SOP CEO pay in countries eventually adopting SOP as well as CEO
pay in countries never adopting SOP) They find that firms in a post-SOP regime (i) exhibit
lower CEO pay levels (but only in countries adopting an advisory SOP vote) a finding driven by
a relative decline in equity awards and limited to CEOs and not the other top executives (ii)
higher pay-performance sensitivity (the authors do not look at positive and negative performance
separately) The current version of the paper does not examine whether these findings are more
pronounced in (or driven by) firms with excess CEO pay before the adoption of SOP and in
firms experiencing adverse SOP votes
The drawback of the research design in Correa and Lel (2013) is that it is not clear whether the
results reflect time-series changes within countries adopting SOP (pre vs post) or difference
between firms in SOP and non-SOP countries For example further tests in the paper suggest
that the lower CEO pay levels for post-SOP observations are not driven by a pay decrease
relative to the pre-SOP period (in fact country-by-country analyses indicate no change in CEO
pay levels after the adoption of SOP) but by the comparison with non-SOP countries Also while
the authors employ reasonable econometric solutions to deal with this issue some concerns
remains as to whether non-SOP countries can be a good benchmark for what would have
happened in SOP countries had SOP not been adopted (in other words if only countries with
problematic CEO practices adopt SOP it is possible that changes in CEO pay reflect factors
leading to the adoption of SOP rather than the effect of SOP laws per se) Notwithstanding these
concerns these findings are intriguing and future versions of this recent study have the potential
to provide more conclusive evidence on the effect of SOP
While in the US SOP was mandated by the Dodd-Frank Act its implementation was left to the
Securities and Exchange Commission (SEC) The final rule issued in January 2011 exempted
lsquosmallrsquo firms (ie firms with a public float below $75 million) from the SOP provision for two
years Iliev and Vitanova (2013) exploit this setting to estimate the effect of SOP on CEO
compensation In particularly using a regression discontinuity design they compare changes in
CEO pay of firms just above the SEC-imposed threshold to changes in CEO pay of firms just
below the threshold (but otherwise quite similar) They find no differences in terms of level and
composition of CEO pay and in terms of golden parachutes concluding that SOP had no impact
on CEO pay While the identification strategy chosen by the authors has the benefit of reducing
endogeneity issues it comes with the price of focusing only on fairly small firms (those around
the $75 million public float threshold) where CEO pay may not be a problem in the first place
(in which case the lack of impact of SOP would not be surprising but would not speak to its
impact at large firms traditionally the target of compensation-related criticism)
Cuntildeat Gine and Guadalupe (2013) also use a regression discontinuity design but in a different
setting In particular they analyze nonbinding shareholder proposals to adopt SOP (250
proposals between 2006 and 2010) Comparing the changes in CEO pay for firms voluntarily
adopting SOP in response to the proposals to firms not adopting SOP would be difficult since
the adoption decision is endogenous Similarly it would be difficult to compare changes in CEO
pay between firms where the proposal is passed (and thus significantly more likely to be
adopted) and firms where the proposal fails to be approved since the voting outcome is
endogenous as well
Hence Cuntildeat Gine and Guadalupe (2013) use instead a regression discontinuity design that
compares firms where the proposals receive slightly more than 50 of the votes (the threshold
for approval) to otherwise similar firms where the proposals receive slightly less than 50
Using this approach they conclude that the passing and adoption of SOP proposals is not
associated with changes in the level of CEO pay nor its composition and structure (mix of cash
and equity equity incentives level of perks and deferred pay) In some sense this finding
complements the one in Iliev and Vitanova (2013) in that it is based on large firms (shareholder
proposals are typically submitted at large SampP 500 firms) However both Cai and Walkling
(2011) and Cuntildeat Gine and Guadalupe (2013) show that firms targeted by SOP proposals do not
exhibit excess CEO pay or weaker governance8 Hence the lack of an effect at these firms may
8 Activists submitting SOP proposals chose to focus on a broad sample of large firms rather than target only firms
with problematic CEO pay practices to obtain visibility and show to policy-makers that investorrsquos support for SOP
was not limited to problematic firms (Ferri and Weber 2009)
not be surprising More generally the findings cannot be necessarily generalized to firms away
from the 50 threshold or to firms not targeted by SOP shareholder proposals
B Effect of SOP votes on compensation practices
Analyses of the effect of SOP on level and composition of CEO pay may not capture a number
of other changes to compensation contracts that may be induced by SOP votes but are not
reflected in measure of CEO pay For example the introduction of performance-based vesting
provisions is typically not reflected in the estimates of fair value of equity grants used by
researchers in measuring CEO pay Many other changes (eg terms of severance packages) will
only be reflected in measures of CEO pay contingent upon the occurrence of certain events
Hence it is important to complement the evidence in section IIA with an analysis of firmsrsquo
disclosures of changes to observable provisions of compensation contracts Another benefit of
this approach is that it captures the changes that boards explicitly present as a response to SOP
votes and thus it acts as a reality check on regression-based inferences For example a
regression-based finding of say a decrease in CEO pay level would be more credible if
supported by evidence that firms disclose a decision to reduce target levels of CEO pay in
response to an adverse SOP vote (an action that presumably they have an incentive to disclose to
get rewarded by shareholders in subsequent votes)
Two studies collect data on specific changes made to compensation contracts explicitly in
response to SOP votes in the UK and the US using firmsrsquo disclosures in their proxy statements
Ferri and Maber (2013) examine the compensation reports of a sample of UK firms before and
after the first SOP vote and find that firms experiencing higher voting dissent were significantly
more likely to remove compensation provisions criticized by investors as lsquorewards for failurersquo
relative to a matched sample of firms experiencing lower dissent For example they report that
among firms with long notice periods (implying larger severance payments ) the percentage of
high dissent firms that shortened them after the vote (800) was significantly higher than before
the vote (200) and also significantly higher than among low dissent firms after the vote
(333) They report similar findings for another controversial practice namely the presence of
retesting provisions in the performance-based vesting conditions of equity grants Most firms
indicate that these changes were the result of consultations with their major institutional
investors Firmsrsquo responsiveness to SOP votes was lsquorewardedrsquo with a substantial increase in
favorable SOP votes at the subsequent annual meeting Ferri and Maber (2013) also find that
many firms experiencing low voting dissent at the first SOP vote had removed these practices
before the vote9 highlighting the importance of accounting for ex ante effects when assessing the
impact of a new regulation
Ertimur Ferri and Oesch (2013) examine the effect of SOP on compensation practices in the US
in 2011 (first proxy season under mandatory SOP) Their key findings are generally similar to
the evidence from the UK (i) failed SOP votes are rare (about 2 of the sample) though cases
of substantial dissent are more frequent (ii) more than 55 of the firms experiencing significant
voting dissent respond by making material changes to their compensation plan during the
subsequent year (eg introduction of performance-based vesting conditions in equity grants use
of tougher performance targets removal of perks and tax gross-ups removal of controversial
provisions from severance contracts etc) usually in consultation with major institutional
9 In their sample 700 of low dissent firms shortened their notice periods during the year before the SOP vote
investors (iii) firms making changes to their compensation plans receive greater voting support
at the following SOP vote
Ertimur Ferri and Oesch (2013) also highlight the strong influence of the recommendations
released by proxy advisors (particularly Institutional Shareholder Services ISS) with a negative
recommendation being associated with 25 more votes against the compensation plan Further
evidence of the significant influence of ISS is the striking discontinuity in the relation between
the extent of SOP voting dissent and firmsrsquo responsiveness to the vote the percentage of firms
making compensation changes in response to SOP votes jumps from 32 to 72 around the
30 voting dissent threshold Why After the 2011 proxy season ISS had indicated that firms
failing to ldquoadequatelyrdquo respond to SOP voting dissent above 30 would receive a negative
recommendation in 2012 on the SOP proposal and on the election of compensation committee
members
Finally similar to the UK experience there is also anecdotal evidence of analogous
compensation changes being made by US firms ahead of the SOP vote to avoid a negative
proxy advisor recommendation and an adverse shareholder vote (Thomas Palmiter and Cotter
2013 Larcker McCall and Ormazabal 2013)
III The Effect of Say on Pay on Firm Value
In this section we examine the empirical studies on the overall effect of SOP on firm value
A Event studies around the adoption of Say on Pay regulations
As mentioned in the Introduction on April 20 2007 the House of Representatives passed a SOP
Bill by a 2-1 margin On the same day then-Senator Barack Obama introduced a companion
bill (S1181) in the Senate (which was then put on hold by the Senate Banking Committee)
While the SOP Billrsquos approval was expected (Democrats were in control of the House and
supported the SOP Bill) the 2-1 margin was unexpected suggesting some support for SOP
among Republicans
Cai and Walkling (2011) examine the market reaction to the Housersquos approval of the SOP Bill
for a sample of firms in the SampP 1500 index and document a positive reaction for firms more
likely to benefit from greater shareholder voice over executive pay namely firms with high
abnormal CEO cash pay (defined as the difference between actual CEO cash pay and the level
predicted based on known economic determinants such a size performance industry) firms
with low pay-for-performance sensitivity firms with a history of shareholders willing to vote
against compensation-related management proposals and firms with a history of responsiveness
to shareholder pressure on compensation issues In other words they find a positive reaction in
firms where the compensation problems are more severe and a say on pay vote is more likely to
trigger a response (eg more shareholders willing to vote against management and greater firmsrsquo
propensity to respond to an adverse vote) Note however that they do not find a significant
impact for firms with high abnormal equity and total CEO pay
Larcker Ormazabal and Taylor (2011) also examine the market reaction to the Housersquos approval
of the SOP Bill (in addition to other regulatory events related to executive pay and other
governance provisions) Similar to Cai and Wakling (2011) they fail to find a significant impact
for firms with high abnormal total CEO pay (they do not examine the price reaction for firms
with abnormal cash CEO pay or firms with lower pay-performance sensitivity)
A concern with both studies is that the Housersquos approval of a bill does not guarantee passage in
the Senate and approval by the White House In fact the press at the time reported that the
prospects of the bill in the Senate were uncertain (New York Times 2007) and the Bush White
House openly opposed the Bill (Associated Press 2007) Hence it is not clear the extent to
which this event increased the likelihood of a say on pay legislation Arguably as noted in
Section I the only lsquoeventrsquo that substantially increased the likelihood of say on pay legislation
was the financial crisis and the related perception of compensation abuses (eg outrage over AIG
bonuses)
Ferri and Maber (2013) argue that the announcement of the submission of SOP regulation to the
Parliament in the United Kingdom in June 2002 offers a more powerful setting for an event
study since it was largely unexpected and increased substantially the probability of SOP
adoption (Parliamentrsquos approval was virtually guaranteed) Using this event they document
positive abnormal returns for firms with excess CEO pay combined with poor performance and
for firms with controversial pay practices that weaken the penalties for poor performance (those
practices were often removed in response to SOP votes as discussed earlier in Section IIB)
They interpret their findings as consistent with shareholders perceiving SOP as a value
enhancing monitoring mechanism for firms with low pay-to-poor-performance sensitivity
Finally a recent study by Iliev and Vitanova (2013) examines the market reaction around the
announcement of the SEC final SOP rule that on January 25 2011 exempted small firms (ie
firms with a public float below $75 million) from adopting SOP for two years (press reports at
the time noted that the exemption could become permanent) An appealing feature of this setting
is that the SEC decision was an unexpected reversal from the rule proposed in October 2010 rule
under which no publicly traded firm would be exempted
Iliev and Vitanova (2013) compare the market reaction to this announcement for firms just above
the SEC-imposed threshold and thus subject to the SOP rule and firms just below the threshold
(but otherwise quite similar) and thus exempted from SOP While they find not significant
abnormal returns for either group the abnormal returns for exempted firms are significantly
more negative a finding that the authors interpret as evidence of a positive value effect of
mandatory SOP While the setting is an interesting one some concerns remain as exempted
firms are fairly small and the magnitude of the executive pay problem (if any) is unlikely to be
large enough to explain the differential market reaction Also the same study finds no effect on
CEO pay level and composition (see Section IIA) Hence the reason for the differential stock
price reaction remains unclear It would be helpful to know if the event study result is driven by
(or stronger for) exempted firms with excessive pay or problematic pay practices
B Event studies around Shareholder Proposals to Adopt Say on Pay
Starting in 2006 shareholder activists led by union pension funds submitted shareholder
proposal to adopt SOP at hundreds of US firms in an attempt to induce voluntary or mandatory
widespread adoption of SOP
Cai and Walkling (2011) examine the market reaction to proxy filings and annual meetings of
113 firms targeted by nonbinding shareholder proposals to adopt SOP between 2006 and 2008
On average they find insignificant returns around both the submission of the proposal (proxy
filing date) and the vote on the proposal (annual meeting date) But they also find that when the
proposals are filed by union pension funds the abnormal returns (still insignificant) are more
negative than when filed by other activists and that when the proposal is defeated the abnormal
returns (while insignificant) are more positive than when the proposal is passed (it is not clear
whether this result is driven by union-sponsored proposals) They interpret these findings as
evidence that the market views SOP proposals filed by union pension funds are driven by special
interests rather than value maximization10
However caution is required in interpreting this
evidence Virtually all activists filing SOP proposals were coordinated by a group of investors
(mostly but not only union pension funds) who made available to other activists a template for
SOP proposals and a list of target firms (Ferri and Weber 2009) Hence the distinction between
union and non-union proponents in the context of SOP proposals may be only apparent More
importantly the list of target firms was publicly available months before the proxy filing dates
so it is not clear whether the proxy statements contained any new information Similarly the
voting outcome of the SOP proposals may largely be anticipated (based on the composition of
institutional owners proxy advisorsrsquo recommendations etc) Finally the analyses do not control
for other (potentially new) information contained in proxy statements (eg executive
compensation report other items up for a vote at the annual meeting) or other events occurring at
the annual meeting (eg shareholder votes on other items)
Cuntildeat Gine and Guadalupe (2013) also examine the market reaction to the voting outcome of
nonbinding shareholder proposals to adopt SOP but to alleviate these concerns they employ a
lsquofuzzyrsquo regression discontinuity design (RDD) essentially comparing the stock price reaction to
SOP proposals that pass by a small margin to the reaction to SOP proposals that fail by a small
10
Consistent with this interpretation the authors also find that SOP proposals generally targeted larger firms rather
than firms with excessive pay or lower pay-performance sensitivity However as noted by Ferri and Sandino (2009)
activists typically submit shareholder proposals aimed at promoting policy reforms at a broad sample of large firms
rather than the firms that would most benefit from the proposals because they believe that showing widespread
support for the proposal across all types of firms enhances its credibility with policy makers
margin (Cuntildeat Gine and Guadalupe 2012 Ertimur Ferri and Oesch forthcoming) The underlying idea
is that firms around the threshold are likely to have similar characteristics (indeed they do as the
authors show) but differ in the likelihood of implementation which is typically much higher for
proposals passing the threshold (Ertimur Ferri and Stubben 2010 Thomas and Cotter 2007)
Indeed the authors show that the likelihood of subsequently adopting SOP is 48 higher when
the SOP proposal passes by a small margin relatively to when it fails by a small margin Besides
because in these close-call situations the voting outcome is uncertain its resolution (pass or fail)
is likely to convey new information about the likelihood of SOP adoption making this setting
suitable to an event study11
Using this RDD approach Cuntildeat Gine and Guadalupe (2013) find that on the day of the vote a
SOP proposal that passes by a small margin yields an abnormal return of 24 relative to one
that fails (after controlling for other proposals voted upon at the same meeting) and estimate the
lsquofullrsquo value of SOP at 46 (after taking into account the increase in the probability of SOP
implementation)12
Aside from the issues discussed in Section IA (ie generalizability to other
firms) this estimate seems too large to reflect the present value of future reductions in excess
CEO pay (as noted by the authors their estimate of the value of SOP is equivalent to the
estimated value of removing two anti-takeover provision based on Cuntildeat Gine and Guadalupe
2012) particularly because (i) the authors find no evidence of subsequent changes in levels and
composition of CEO pay for firms implementing SOP (see Section IIA) an (ii) the sample firms
11
Consistent with this observation the authors find no market reaction to the voting outcome of SOP shareholder
proposals away from the threshold
12 Since the outcome of the vote is not binding the 24 market reaction only reflects the expected increase in the
probability of SOP adoption and thus understates the value of the SOP provision
do not appear to be characterized by excess CEO pay However the authors also find significant
improvements in subsequent operating performance and efficiency as a result of passing SOP
proposals and thus conjecture that the SOP vote is viewed as a tool to express a vote of
confidence in management performance more than a way to change compensation practices and
that the large positive market reaction reflects expected performance improvements under this
tighter monitoring regime While this is an interesting idea it remains unclear what additional
ldquoteethrdquo a SOP vote may provide over existing (non-compensation related) monitoring
mechanisms shareholders can express (and do often express) their dissatisfaction with
management performance when voting on director elections (Cai Garner and Walkling 2009)
C Event studies around compensation changes induced by SOP
Two studies examine announcements of changes to compensation plans related to SOP Larcker
McCall and Ormazabal (2013) examine the market reaction to (non-contaminated) compensation
changes disclosed in 8-K filings by Russell 3000 firms in the US during the year before the first
SOP vote They find that changes that appear to be made to avoid a negative proxy advisorrsquos
recommendation and thus a negative SOP vote are associated with a negative abnormal return
of -044 whereas other compensation changes are not associated with a significant stock price
reaction Ertimur Ferri and Oesch (2013) perform a similar analysis but focusing on
compensation changes made after the first SOP vote (and explicitly in response to the vote) by
firms receiving a negative recommendation and a high voting dissent in 2011 They fail to find
significant abnormal returns even for the subset of compensation changes that resulted in a
positive recommendation and low dissent in 2012 (and thus were presumably perceived to be
adequate and material by proxy advisors and voting shareholders)
D Other approaches effect of SOP on Tobinrsquos Q
In their cross-country study Correa and Lel (2013) also examine the effect of SOP laws on
Tobinrsquos Q and report a 36 increase in firm value following the adoption of the SOP laws
(more precisely a 36 higher firm value for post-SOP observations relative to a control sample
that pools together pre-SOP and non-SOP observations) They acknowledge that this increase in
firm value is too large to be justified by the relative decrease in CEO pay that they document and
suggest (but do not test) that it may reflect better alignment of pay and performance13
A key
concern is that higher valuation in countries with SOP laws may reflect other governance
changes introduced at the same time While the study controls for other compensation-related
laws there may be other non-compensation regulations introduced with SOP and with a
potentially larger impact on firm value
IV What have we learned
Adoption of SOP in the US has been long advocated by those who contend that executive pay is
a manifestation of rather than a solution to the agency problem and by those who favor great
shareholder involvement in corporate decisions A growing body of research is starting to
investigate the effect of SOP on executive pay and firm value It is premature to draw definitive
conclusions also because these studies differ in methodologies and settings but three points
seem to emerge First there is robust evidence that boards respond to SOP votes when
13
They also find that the firm value increase is higher when the relative decrease in CEO pay results in a lower pay
differential between CEO and other top managers (CEO pay slice) implying that a source of value creation may be
the reduced pay inequality among the top management team But this seems unlikely to explain the change in
Tobinrsquos Q since the increase occurs for both firms with advisory SOP laws and firms with binding SOP laws and
there is no change in CEO pay slice in firms subject to binding SOP laws
shareholders use it ldquovoicerdquo is heardrdquo Both in the US and UK studies show that not only firms
failing to win the SOP vote but also firms facing substantial dissent (20-30 of the votes
against) make changes to their compensation plans in consultation with institutional investors
and proxy advisors Second in both the US and UK institutional investors appear to use the
power of SOP votes to pressure firms into strengthening the link (or the perceived link) between
pay and performance (particularly on the downside) but not to pressure firms to reduce target
levels of pay suggesting a reluctance of institutional investors to lsquoregulatersquo executive pay Third
and consistent with the above most studies examining the aggregate effect of SOP on CEO pay
find some increase in pay-performance sensitivity but not effect on the level and growth of CEO
pay While some observers may interpret the lack of an effect on CEO pay levels as evidence of
the failure of SOP one should note that SOP per se is a neutral tool Its impact depends on what
investors choose to lsquosay on payrsquo Fourth based on my interpretation of the existing research to
date there is little evidence of an effect of SOP on firm value The studies documenting a
positive price impact appear subject to alternative interpretations or not plausible in their
findings (eg they fail to find an effect on CEO pay or the effect is too small to justify the
documented price impact hence it remains unclear what the source of value creation is) Also
there seems to be no stock price reaction to the SOP-induced compensation changes Finally
given the nature of these SOP-induced changes while some of them may be beneficial it seems
hard to believe that they will have a statistically detectable impact on firm value except perhaps
in a handful of firms where the compensation plan is subject to a complete overhaul
Perhaps one explanation for the weak impact of SOP is that executive pay problems were
overstated or that by the time SOP was introduced (more than a decade after the Enron-type
scandals that led to calls for greater shareholder voice) they had been already addressed via
other mechanisms (hedge fund activism monitoring by institutional investors vote-no
campaigns against compensation committee members SEC-mandated pay disclosures)
Overall though based on the evidence to date it does not appear that SOP had a major impact14
Was the adoption of SOP an lsquooptimalrsquo choice from a social welfare point of view This is a
difficult question to answer given the challenges of identifying and measuring all the potential
costs and benefits associated with SOP (likely to change over time and differ across countries)
Perhaps the most positive view of SOP is that its introduction prevented the adoption of other
more radical and intrusive regulatory measures (eg CEO pay caps) In that sense it may be
viewed as an lsquooptimalrsquo answer (ie the lesser of two evils) to the political pressure to reform
executive pay during the financial crisis
14
This interpretation of the evidence is also consistent with the prediction of analytical studies Ozbas and
Matsusaka (2013) model the benefits and costs of shareholder empowerment distinguishing between the right to
approve and the right to propose They show that permitting shareholders to propose directors or policies can cause
value-maximizing managers to take value-reducing actions to accommodate activist investors with non-value-
maximizing goals As for the right to approve it is weakly beneficial that is approval rights (such as a SOP vote)
are generally beneficial for shareholders in that they limit the managerrsquos ability to pursue private benefits at
shareholder expense but they have minimal impact on managerial actions and firm value (because the manager in
effect can threaten shareholders with an undesirable status quo if they do not approve the managerrsquos proposed
action)
References
Associated Press 2007 Administration opposes say on pay bill April 19
Bainbridge S 2008 Remarks on say on pay An unjustified incursion on director authority
UCLA School of Law Research Paper No 08-06
Bebchuk L 2005 The Case for Increasing Shareholder Power Harvard Law Review 118833-
917
Bebchuk L 2007 Written testimony submitted before the Committee on Financial Services
United States House of Representatives Hearing on Empowering Shareholders on Executive
Compensation March 8
Bebchuk LA A Cohen and A Ferrell 2009 What Matters in Corporate Governance Review
of Financial Studies 22783-827
Cai J J Garner and R Walkling 2009 Electing Directors Journal of Finance 642387-2419
Cai J and R Walkling 2011 Shareholdersrsquo Say on Pay Does it Create Value Journal of
Financial and Quantitative Analysis 46299ndash339
Coates IV JC (2009) Testimony before the Subcommittee on Securities Insurance and
Investment of the Committee on Banking Housing and Urban Affairs United States Senate
July 29 2009 httppapersssrncomsol3paperscfmabstract_id=1475355
Correa R and U Lel 2013 Say On Pay Laws Executive Compensation CEO Pay Slice and
Firm Value Around the World Federal Reserve Board Working Paper
Cuntildeat V M Gine and M Guadalupe 2012 The Vote is Cast The Effect of Corporate
Governance on Shareholder Value Journal of Finance 671943-1977
Cuntildeat V M Gine and M Guadalupe 2013 Say Pays Shareholder Voice and Firm
Performance ECGI - Finance Working Paper No 373
Del Guercio D L Seery and T Woidtke 2008 Do Boards Pay Attention When Institutional
Investors lsquoJust Vote Norsquo Journal of Financial Economics 9084-103
Ertimur Y F Ferri and S Stubben 2010 Board of Directorslsquo Responsiveness to Shareholders
Evidence from Shareholder Proposals Journal of Corporate Finance 1653-72
Ertimur Y F Ferri and V Muslu 2011 Shareholder Activism and CEO Pay Review of
Financial Studies 24535ndash592
Ertimur Y F Ferri and D Oesch 2013 Shareholder Votes and Proxy Advisors Evidence from
Say on Pay Journal of Accounting Research 51951-996
Ertimur Y F Ferri and D Oesch forthcoming Does the Director Election System Matter
Evidence from Majority Voting Review of Accounting Studies forthcoming
Ferri F 2012 Low-Cost Shareholder Activism A Review of the Evidence Research
Handbook on the Economics of Corporate Law ed CA Hill and BH McDonnell
Edward Elgar Publishing
Ferri F and D Maber 2013 Say on Pay Votes and CEO Compensation Evidence from the UK
Review of Finance 17527-563
Ferri F and D Oesch 2013 Management Influence on Investors Evidence from Shareholder
Votes on the Frequency of Say on Pay Columbia Business School Research Paper No 13-17
Ferri F and T Sandino 2009 The impact of shareholder activism on financial reporting and
compensation The case of employee stock options expensing The Accounting Review 84433ndash
466
Ferri F and J Weber 2009 AFSCME vs Mozilohellipand Say on Pay for All (A) Harvard
Business School Case 109-009
Gompers P J Ishii and A Metrick 2003 Corporate Governance and Equity Prices Quarterly
Journal of Economics 118107-155
Gordon J 2009 Say on Paylsquo Cautionary Notes on the UK Experience and the Case for
Shareholders Opt-In Harvard Journal on Legislation 46323-368
Iliev P and S Vitanova 2013 The effect of the Say on Pay in the US Pennsylvania State
University Working Paper
Kaplan S 2007 Testimony of Steven N Kaplan on Empowering Shareholders on Executive
Compensation and HR 1257 the Shareholder Vote on Executive Compensation Act before the
Committee on Financial Services United States House of Representatives March 8
Larcker D A McCall and G Ormazabal 2013 The Economic Consequences of Proxy
Advisor Say-on-Pay Voting Policies Working Paper Stanford University
Larcker D G Ormazabal and D Taylor 2011 The market reaction to corporate governance
regulation Journal of Financial Economics 101431ndash448
Levit D and N Malenko 2011 Non-binding voting for shareholder proposals Journal of
Finance 661579ndash1614
Mishel L and N Sabadish 2012 How executive compensation and financial-sector pay have
fueled income inequality Issue Brief 331 Economic Policy Institute
Murphy K 2012 The Politics of Pay A Legislative History of Executive Compensation in
Jennifer Hill and Randall Thomas eds Research Handbook on Executive Pay Edward Elgar
Publishers
New York Times 2007 House votes to give investors say on executive pay April 21
Norris F 2004 Corporate democracy and the power to embarrass New York Times March 4
Ozbas O and J Matsusaka2013 Managerial Accommodation Proxy Access and the Cost of
Shareholder Empowerment University of Southern California CLEO Research Paper Series No
C12-1
Thomas R and J Cotter 2007 Shareholder proposals in the new millennium Shareholder
support board response and market reaction Journal of Corporate Finance 13 368ndash391
Thomas R A Palmiter and J Cotter 2012 Dodd-Franks Say on Pay Will it Lead to a Greater
Role for Shareholders in Corporate Governance Cornell Law Review 971213-1266
Thomas R and C Van der Elst 2013 The International Scope of Say on Pay ECGI Law
Working Paper No227
Africa and Switzerland adopted some version of a binding SOP vote (on the future compensation
policy) which is now also being considered by the European Union7
When putting SOP into its historical context it is also important to appreciate that SOP has come
to epitomize a broader movement toward greater shareholder democracy beyond the issue of
executive pay For example during the same period shareholder activists have long
(unsuccessfully) lobbied for a proxy access rule that would have made it easier for shareholders
to oust corporate directors and nominate their own candidates Hence many observers often
view the successes and failures of SOP as speaking to the potential effects of other reforms
aimed at increasing shareholder power making the academic research on SOP all more relevant
II The Effect of Say on Pay on Executive Compensation
A Effect of SOP votes on level and composition of executive pay
As discussed earlier United Kingdom was the first country to adopt SOP in 2002 for publicly
traded firms Ferri and Maber (2013) report that in most cases shareholders voted in favor of
compensation plans Failed SOP votes (ie greater than 50 of votes against) were rare (2 of
the sample) though highly publicized in the press However about one-fourth of the sample
firms received more than 20 of votes against
To capture the overall effect of SOP on CEO pay Ferri and Maber (2013) examine the
sensitivity of CEO pay to its economic determinants over the 2000-2005 period (ie before and
after the introduction of SOP regulation) and document a significant increase in the sensitivity of
7 Some countries have already modified their first SOP legislation In 2013 the UK added to the non-binding
backward-looking SOP vote on the current compensation report a binding forward-looking vote on the proposed
compensation policy In 2012 Australia modified the non-binding SOP vote to introduce the so-called ldquotwo-strikesrdquo
rule Under the rule if 25 of shareholders vote against a companyrsquos remuneration report at two consecutive annual
general meetings the entire board may have to stand for re-election within three months For more details about the
international development of SOP see Thomas and Van der Elst (2013)
CEO pay to poor performance They also find that this increase does not occur for a subset of
UK firms exempted from the SOP regulation (firms traded on the Alternative Investment Market
a sub-market of the London Stock Exchange with a more flexible regulatory system) suggesting
that the result reflects the impact of SOP rather than a general trend affecting all firms Besides
Ferri and Maber (2013) find that the increase is more pronounced in firms experiencing high
voting dissent and firms with high abnormal CEO pay before the adoption of SOP consistent
with a causal impact of SOP In contrast they fail to find any effect of SOP on the level of pay
Correa and Lel (2013) examine the effect of SOP laws on CEO pay using a large cross-country
sample of about 103000 firm-year observations from 39 countries including 12 countries that
adopted some (advisory or binding) version of SOP They essentially compare post-SOP CEO
pay at firms of countries adopting SOP to a control sample which includes all non-SOP
observations (that is pre-SOP CEO pay in countries eventually adopting SOP as well as CEO
pay in countries never adopting SOP) They find that firms in a post-SOP regime (i) exhibit
lower CEO pay levels (but only in countries adopting an advisory SOP vote) a finding driven by
a relative decline in equity awards and limited to CEOs and not the other top executives (ii)
higher pay-performance sensitivity (the authors do not look at positive and negative performance
separately) The current version of the paper does not examine whether these findings are more
pronounced in (or driven by) firms with excess CEO pay before the adoption of SOP and in
firms experiencing adverse SOP votes
The drawback of the research design in Correa and Lel (2013) is that it is not clear whether the
results reflect time-series changes within countries adopting SOP (pre vs post) or difference
between firms in SOP and non-SOP countries For example further tests in the paper suggest
that the lower CEO pay levels for post-SOP observations are not driven by a pay decrease
relative to the pre-SOP period (in fact country-by-country analyses indicate no change in CEO
pay levels after the adoption of SOP) but by the comparison with non-SOP countries Also while
the authors employ reasonable econometric solutions to deal with this issue some concerns
remains as to whether non-SOP countries can be a good benchmark for what would have
happened in SOP countries had SOP not been adopted (in other words if only countries with
problematic CEO practices adopt SOP it is possible that changes in CEO pay reflect factors
leading to the adoption of SOP rather than the effect of SOP laws per se) Notwithstanding these
concerns these findings are intriguing and future versions of this recent study have the potential
to provide more conclusive evidence on the effect of SOP
While in the US SOP was mandated by the Dodd-Frank Act its implementation was left to the
Securities and Exchange Commission (SEC) The final rule issued in January 2011 exempted
lsquosmallrsquo firms (ie firms with a public float below $75 million) from the SOP provision for two
years Iliev and Vitanova (2013) exploit this setting to estimate the effect of SOP on CEO
compensation In particularly using a regression discontinuity design they compare changes in
CEO pay of firms just above the SEC-imposed threshold to changes in CEO pay of firms just
below the threshold (but otherwise quite similar) They find no differences in terms of level and
composition of CEO pay and in terms of golden parachutes concluding that SOP had no impact
on CEO pay While the identification strategy chosen by the authors has the benefit of reducing
endogeneity issues it comes with the price of focusing only on fairly small firms (those around
the $75 million public float threshold) where CEO pay may not be a problem in the first place
(in which case the lack of impact of SOP would not be surprising but would not speak to its
impact at large firms traditionally the target of compensation-related criticism)
Cuntildeat Gine and Guadalupe (2013) also use a regression discontinuity design but in a different
setting In particular they analyze nonbinding shareholder proposals to adopt SOP (250
proposals between 2006 and 2010) Comparing the changes in CEO pay for firms voluntarily
adopting SOP in response to the proposals to firms not adopting SOP would be difficult since
the adoption decision is endogenous Similarly it would be difficult to compare changes in CEO
pay between firms where the proposal is passed (and thus significantly more likely to be
adopted) and firms where the proposal fails to be approved since the voting outcome is
endogenous as well
Hence Cuntildeat Gine and Guadalupe (2013) use instead a regression discontinuity design that
compares firms where the proposals receive slightly more than 50 of the votes (the threshold
for approval) to otherwise similar firms where the proposals receive slightly less than 50
Using this approach they conclude that the passing and adoption of SOP proposals is not
associated with changes in the level of CEO pay nor its composition and structure (mix of cash
and equity equity incentives level of perks and deferred pay) In some sense this finding
complements the one in Iliev and Vitanova (2013) in that it is based on large firms (shareholder
proposals are typically submitted at large SampP 500 firms) However both Cai and Walkling
(2011) and Cuntildeat Gine and Guadalupe (2013) show that firms targeted by SOP proposals do not
exhibit excess CEO pay or weaker governance8 Hence the lack of an effect at these firms may
8 Activists submitting SOP proposals chose to focus on a broad sample of large firms rather than target only firms
with problematic CEO pay practices to obtain visibility and show to policy-makers that investorrsquos support for SOP
was not limited to problematic firms (Ferri and Weber 2009)
not be surprising More generally the findings cannot be necessarily generalized to firms away
from the 50 threshold or to firms not targeted by SOP shareholder proposals
B Effect of SOP votes on compensation practices
Analyses of the effect of SOP on level and composition of CEO pay may not capture a number
of other changes to compensation contracts that may be induced by SOP votes but are not
reflected in measure of CEO pay For example the introduction of performance-based vesting
provisions is typically not reflected in the estimates of fair value of equity grants used by
researchers in measuring CEO pay Many other changes (eg terms of severance packages) will
only be reflected in measures of CEO pay contingent upon the occurrence of certain events
Hence it is important to complement the evidence in section IIA with an analysis of firmsrsquo
disclosures of changes to observable provisions of compensation contracts Another benefit of
this approach is that it captures the changes that boards explicitly present as a response to SOP
votes and thus it acts as a reality check on regression-based inferences For example a
regression-based finding of say a decrease in CEO pay level would be more credible if
supported by evidence that firms disclose a decision to reduce target levels of CEO pay in
response to an adverse SOP vote (an action that presumably they have an incentive to disclose to
get rewarded by shareholders in subsequent votes)
Two studies collect data on specific changes made to compensation contracts explicitly in
response to SOP votes in the UK and the US using firmsrsquo disclosures in their proxy statements
Ferri and Maber (2013) examine the compensation reports of a sample of UK firms before and
after the first SOP vote and find that firms experiencing higher voting dissent were significantly
more likely to remove compensation provisions criticized by investors as lsquorewards for failurersquo
relative to a matched sample of firms experiencing lower dissent For example they report that
among firms with long notice periods (implying larger severance payments ) the percentage of
high dissent firms that shortened them after the vote (800) was significantly higher than before
the vote (200) and also significantly higher than among low dissent firms after the vote
(333) They report similar findings for another controversial practice namely the presence of
retesting provisions in the performance-based vesting conditions of equity grants Most firms
indicate that these changes were the result of consultations with their major institutional
investors Firmsrsquo responsiveness to SOP votes was lsquorewardedrsquo with a substantial increase in
favorable SOP votes at the subsequent annual meeting Ferri and Maber (2013) also find that
many firms experiencing low voting dissent at the first SOP vote had removed these practices
before the vote9 highlighting the importance of accounting for ex ante effects when assessing the
impact of a new regulation
Ertimur Ferri and Oesch (2013) examine the effect of SOP on compensation practices in the US
in 2011 (first proxy season under mandatory SOP) Their key findings are generally similar to
the evidence from the UK (i) failed SOP votes are rare (about 2 of the sample) though cases
of substantial dissent are more frequent (ii) more than 55 of the firms experiencing significant
voting dissent respond by making material changes to their compensation plan during the
subsequent year (eg introduction of performance-based vesting conditions in equity grants use
of tougher performance targets removal of perks and tax gross-ups removal of controversial
provisions from severance contracts etc) usually in consultation with major institutional
9 In their sample 700 of low dissent firms shortened their notice periods during the year before the SOP vote
investors (iii) firms making changes to their compensation plans receive greater voting support
at the following SOP vote
Ertimur Ferri and Oesch (2013) also highlight the strong influence of the recommendations
released by proxy advisors (particularly Institutional Shareholder Services ISS) with a negative
recommendation being associated with 25 more votes against the compensation plan Further
evidence of the significant influence of ISS is the striking discontinuity in the relation between
the extent of SOP voting dissent and firmsrsquo responsiveness to the vote the percentage of firms
making compensation changes in response to SOP votes jumps from 32 to 72 around the
30 voting dissent threshold Why After the 2011 proxy season ISS had indicated that firms
failing to ldquoadequatelyrdquo respond to SOP voting dissent above 30 would receive a negative
recommendation in 2012 on the SOP proposal and on the election of compensation committee
members
Finally similar to the UK experience there is also anecdotal evidence of analogous
compensation changes being made by US firms ahead of the SOP vote to avoid a negative
proxy advisor recommendation and an adverse shareholder vote (Thomas Palmiter and Cotter
2013 Larcker McCall and Ormazabal 2013)
III The Effect of Say on Pay on Firm Value
In this section we examine the empirical studies on the overall effect of SOP on firm value
A Event studies around the adoption of Say on Pay regulations
As mentioned in the Introduction on April 20 2007 the House of Representatives passed a SOP
Bill by a 2-1 margin On the same day then-Senator Barack Obama introduced a companion
bill (S1181) in the Senate (which was then put on hold by the Senate Banking Committee)
While the SOP Billrsquos approval was expected (Democrats were in control of the House and
supported the SOP Bill) the 2-1 margin was unexpected suggesting some support for SOP
among Republicans
Cai and Walkling (2011) examine the market reaction to the Housersquos approval of the SOP Bill
for a sample of firms in the SampP 1500 index and document a positive reaction for firms more
likely to benefit from greater shareholder voice over executive pay namely firms with high
abnormal CEO cash pay (defined as the difference between actual CEO cash pay and the level
predicted based on known economic determinants such a size performance industry) firms
with low pay-for-performance sensitivity firms with a history of shareholders willing to vote
against compensation-related management proposals and firms with a history of responsiveness
to shareholder pressure on compensation issues In other words they find a positive reaction in
firms where the compensation problems are more severe and a say on pay vote is more likely to
trigger a response (eg more shareholders willing to vote against management and greater firmsrsquo
propensity to respond to an adverse vote) Note however that they do not find a significant
impact for firms with high abnormal equity and total CEO pay
Larcker Ormazabal and Taylor (2011) also examine the market reaction to the Housersquos approval
of the SOP Bill (in addition to other regulatory events related to executive pay and other
governance provisions) Similar to Cai and Wakling (2011) they fail to find a significant impact
for firms with high abnormal total CEO pay (they do not examine the price reaction for firms
with abnormal cash CEO pay or firms with lower pay-performance sensitivity)
A concern with both studies is that the Housersquos approval of a bill does not guarantee passage in
the Senate and approval by the White House In fact the press at the time reported that the
prospects of the bill in the Senate were uncertain (New York Times 2007) and the Bush White
House openly opposed the Bill (Associated Press 2007) Hence it is not clear the extent to
which this event increased the likelihood of a say on pay legislation Arguably as noted in
Section I the only lsquoeventrsquo that substantially increased the likelihood of say on pay legislation
was the financial crisis and the related perception of compensation abuses (eg outrage over AIG
bonuses)
Ferri and Maber (2013) argue that the announcement of the submission of SOP regulation to the
Parliament in the United Kingdom in June 2002 offers a more powerful setting for an event
study since it was largely unexpected and increased substantially the probability of SOP
adoption (Parliamentrsquos approval was virtually guaranteed) Using this event they document
positive abnormal returns for firms with excess CEO pay combined with poor performance and
for firms with controversial pay practices that weaken the penalties for poor performance (those
practices were often removed in response to SOP votes as discussed earlier in Section IIB)
They interpret their findings as consistent with shareholders perceiving SOP as a value
enhancing monitoring mechanism for firms with low pay-to-poor-performance sensitivity
Finally a recent study by Iliev and Vitanova (2013) examines the market reaction around the
announcement of the SEC final SOP rule that on January 25 2011 exempted small firms (ie
firms with a public float below $75 million) from adopting SOP for two years (press reports at
the time noted that the exemption could become permanent) An appealing feature of this setting
is that the SEC decision was an unexpected reversal from the rule proposed in October 2010 rule
under which no publicly traded firm would be exempted
Iliev and Vitanova (2013) compare the market reaction to this announcement for firms just above
the SEC-imposed threshold and thus subject to the SOP rule and firms just below the threshold
(but otherwise quite similar) and thus exempted from SOP While they find not significant
abnormal returns for either group the abnormal returns for exempted firms are significantly
more negative a finding that the authors interpret as evidence of a positive value effect of
mandatory SOP While the setting is an interesting one some concerns remain as exempted
firms are fairly small and the magnitude of the executive pay problem (if any) is unlikely to be
large enough to explain the differential market reaction Also the same study finds no effect on
CEO pay level and composition (see Section IIA) Hence the reason for the differential stock
price reaction remains unclear It would be helpful to know if the event study result is driven by
(or stronger for) exempted firms with excessive pay or problematic pay practices
B Event studies around Shareholder Proposals to Adopt Say on Pay
Starting in 2006 shareholder activists led by union pension funds submitted shareholder
proposal to adopt SOP at hundreds of US firms in an attempt to induce voluntary or mandatory
widespread adoption of SOP
Cai and Walkling (2011) examine the market reaction to proxy filings and annual meetings of
113 firms targeted by nonbinding shareholder proposals to adopt SOP between 2006 and 2008
On average they find insignificant returns around both the submission of the proposal (proxy
filing date) and the vote on the proposal (annual meeting date) But they also find that when the
proposals are filed by union pension funds the abnormal returns (still insignificant) are more
negative than when filed by other activists and that when the proposal is defeated the abnormal
returns (while insignificant) are more positive than when the proposal is passed (it is not clear
whether this result is driven by union-sponsored proposals) They interpret these findings as
evidence that the market views SOP proposals filed by union pension funds are driven by special
interests rather than value maximization10
However caution is required in interpreting this
evidence Virtually all activists filing SOP proposals were coordinated by a group of investors
(mostly but not only union pension funds) who made available to other activists a template for
SOP proposals and a list of target firms (Ferri and Weber 2009) Hence the distinction between
union and non-union proponents in the context of SOP proposals may be only apparent More
importantly the list of target firms was publicly available months before the proxy filing dates
so it is not clear whether the proxy statements contained any new information Similarly the
voting outcome of the SOP proposals may largely be anticipated (based on the composition of
institutional owners proxy advisorsrsquo recommendations etc) Finally the analyses do not control
for other (potentially new) information contained in proxy statements (eg executive
compensation report other items up for a vote at the annual meeting) or other events occurring at
the annual meeting (eg shareholder votes on other items)
Cuntildeat Gine and Guadalupe (2013) also examine the market reaction to the voting outcome of
nonbinding shareholder proposals to adopt SOP but to alleviate these concerns they employ a
lsquofuzzyrsquo regression discontinuity design (RDD) essentially comparing the stock price reaction to
SOP proposals that pass by a small margin to the reaction to SOP proposals that fail by a small
10
Consistent with this interpretation the authors also find that SOP proposals generally targeted larger firms rather
than firms with excessive pay or lower pay-performance sensitivity However as noted by Ferri and Sandino (2009)
activists typically submit shareholder proposals aimed at promoting policy reforms at a broad sample of large firms
rather than the firms that would most benefit from the proposals because they believe that showing widespread
support for the proposal across all types of firms enhances its credibility with policy makers
margin (Cuntildeat Gine and Guadalupe 2012 Ertimur Ferri and Oesch forthcoming) The underlying idea
is that firms around the threshold are likely to have similar characteristics (indeed they do as the
authors show) but differ in the likelihood of implementation which is typically much higher for
proposals passing the threshold (Ertimur Ferri and Stubben 2010 Thomas and Cotter 2007)
Indeed the authors show that the likelihood of subsequently adopting SOP is 48 higher when
the SOP proposal passes by a small margin relatively to when it fails by a small margin Besides
because in these close-call situations the voting outcome is uncertain its resolution (pass or fail)
is likely to convey new information about the likelihood of SOP adoption making this setting
suitable to an event study11
Using this RDD approach Cuntildeat Gine and Guadalupe (2013) find that on the day of the vote a
SOP proposal that passes by a small margin yields an abnormal return of 24 relative to one
that fails (after controlling for other proposals voted upon at the same meeting) and estimate the
lsquofullrsquo value of SOP at 46 (after taking into account the increase in the probability of SOP
implementation)12
Aside from the issues discussed in Section IA (ie generalizability to other
firms) this estimate seems too large to reflect the present value of future reductions in excess
CEO pay (as noted by the authors their estimate of the value of SOP is equivalent to the
estimated value of removing two anti-takeover provision based on Cuntildeat Gine and Guadalupe
2012) particularly because (i) the authors find no evidence of subsequent changes in levels and
composition of CEO pay for firms implementing SOP (see Section IIA) an (ii) the sample firms
11
Consistent with this observation the authors find no market reaction to the voting outcome of SOP shareholder
proposals away from the threshold
12 Since the outcome of the vote is not binding the 24 market reaction only reflects the expected increase in the
probability of SOP adoption and thus understates the value of the SOP provision
do not appear to be characterized by excess CEO pay However the authors also find significant
improvements in subsequent operating performance and efficiency as a result of passing SOP
proposals and thus conjecture that the SOP vote is viewed as a tool to express a vote of
confidence in management performance more than a way to change compensation practices and
that the large positive market reaction reflects expected performance improvements under this
tighter monitoring regime While this is an interesting idea it remains unclear what additional
ldquoteethrdquo a SOP vote may provide over existing (non-compensation related) monitoring
mechanisms shareholders can express (and do often express) their dissatisfaction with
management performance when voting on director elections (Cai Garner and Walkling 2009)
C Event studies around compensation changes induced by SOP
Two studies examine announcements of changes to compensation plans related to SOP Larcker
McCall and Ormazabal (2013) examine the market reaction to (non-contaminated) compensation
changes disclosed in 8-K filings by Russell 3000 firms in the US during the year before the first
SOP vote They find that changes that appear to be made to avoid a negative proxy advisorrsquos
recommendation and thus a negative SOP vote are associated with a negative abnormal return
of -044 whereas other compensation changes are not associated with a significant stock price
reaction Ertimur Ferri and Oesch (2013) perform a similar analysis but focusing on
compensation changes made after the first SOP vote (and explicitly in response to the vote) by
firms receiving a negative recommendation and a high voting dissent in 2011 They fail to find
significant abnormal returns even for the subset of compensation changes that resulted in a
positive recommendation and low dissent in 2012 (and thus were presumably perceived to be
adequate and material by proxy advisors and voting shareholders)
D Other approaches effect of SOP on Tobinrsquos Q
In their cross-country study Correa and Lel (2013) also examine the effect of SOP laws on
Tobinrsquos Q and report a 36 increase in firm value following the adoption of the SOP laws
(more precisely a 36 higher firm value for post-SOP observations relative to a control sample
that pools together pre-SOP and non-SOP observations) They acknowledge that this increase in
firm value is too large to be justified by the relative decrease in CEO pay that they document and
suggest (but do not test) that it may reflect better alignment of pay and performance13
A key
concern is that higher valuation in countries with SOP laws may reflect other governance
changes introduced at the same time While the study controls for other compensation-related
laws there may be other non-compensation regulations introduced with SOP and with a
potentially larger impact on firm value
IV What have we learned
Adoption of SOP in the US has been long advocated by those who contend that executive pay is
a manifestation of rather than a solution to the agency problem and by those who favor great
shareholder involvement in corporate decisions A growing body of research is starting to
investigate the effect of SOP on executive pay and firm value It is premature to draw definitive
conclusions also because these studies differ in methodologies and settings but three points
seem to emerge First there is robust evidence that boards respond to SOP votes when
13
They also find that the firm value increase is higher when the relative decrease in CEO pay results in a lower pay
differential between CEO and other top managers (CEO pay slice) implying that a source of value creation may be
the reduced pay inequality among the top management team But this seems unlikely to explain the change in
Tobinrsquos Q since the increase occurs for both firms with advisory SOP laws and firms with binding SOP laws and
there is no change in CEO pay slice in firms subject to binding SOP laws
shareholders use it ldquovoicerdquo is heardrdquo Both in the US and UK studies show that not only firms
failing to win the SOP vote but also firms facing substantial dissent (20-30 of the votes
against) make changes to their compensation plans in consultation with institutional investors
and proxy advisors Second in both the US and UK institutional investors appear to use the
power of SOP votes to pressure firms into strengthening the link (or the perceived link) between
pay and performance (particularly on the downside) but not to pressure firms to reduce target
levels of pay suggesting a reluctance of institutional investors to lsquoregulatersquo executive pay Third
and consistent with the above most studies examining the aggregate effect of SOP on CEO pay
find some increase in pay-performance sensitivity but not effect on the level and growth of CEO
pay While some observers may interpret the lack of an effect on CEO pay levels as evidence of
the failure of SOP one should note that SOP per se is a neutral tool Its impact depends on what
investors choose to lsquosay on payrsquo Fourth based on my interpretation of the existing research to
date there is little evidence of an effect of SOP on firm value The studies documenting a
positive price impact appear subject to alternative interpretations or not plausible in their
findings (eg they fail to find an effect on CEO pay or the effect is too small to justify the
documented price impact hence it remains unclear what the source of value creation is) Also
there seems to be no stock price reaction to the SOP-induced compensation changes Finally
given the nature of these SOP-induced changes while some of them may be beneficial it seems
hard to believe that they will have a statistically detectable impact on firm value except perhaps
in a handful of firms where the compensation plan is subject to a complete overhaul
Perhaps one explanation for the weak impact of SOP is that executive pay problems were
overstated or that by the time SOP was introduced (more than a decade after the Enron-type
scandals that led to calls for greater shareholder voice) they had been already addressed via
other mechanisms (hedge fund activism monitoring by institutional investors vote-no
campaigns against compensation committee members SEC-mandated pay disclosures)
Overall though based on the evidence to date it does not appear that SOP had a major impact14
Was the adoption of SOP an lsquooptimalrsquo choice from a social welfare point of view This is a
difficult question to answer given the challenges of identifying and measuring all the potential
costs and benefits associated with SOP (likely to change over time and differ across countries)
Perhaps the most positive view of SOP is that its introduction prevented the adoption of other
more radical and intrusive regulatory measures (eg CEO pay caps) In that sense it may be
viewed as an lsquooptimalrsquo answer (ie the lesser of two evils) to the political pressure to reform
executive pay during the financial crisis
14
This interpretation of the evidence is also consistent with the prediction of analytical studies Ozbas and
Matsusaka (2013) model the benefits and costs of shareholder empowerment distinguishing between the right to
approve and the right to propose They show that permitting shareholders to propose directors or policies can cause
value-maximizing managers to take value-reducing actions to accommodate activist investors with non-value-
maximizing goals As for the right to approve it is weakly beneficial that is approval rights (such as a SOP vote)
are generally beneficial for shareholders in that they limit the managerrsquos ability to pursue private benefits at
shareholder expense but they have minimal impact on managerial actions and firm value (because the manager in
effect can threaten shareholders with an undesirable status quo if they do not approve the managerrsquos proposed
action)
References
Associated Press 2007 Administration opposes say on pay bill April 19
Bainbridge S 2008 Remarks on say on pay An unjustified incursion on director authority
UCLA School of Law Research Paper No 08-06
Bebchuk L 2005 The Case for Increasing Shareholder Power Harvard Law Review 118833-
917
Bebchuk L 2007 Written testimony submitted before the Committee on Financial Services
United States House of Representatives Hearing on Empowering Shareholders on Executive
Compensation March 8
Bebchuk LA A Cohen and A Ferrell 2009 What Matters in Corporate Governance Review
of Financial Studies 22783-827
Cai J J Garner and R Walkling 2009 Electing Directors Journal of Finance 642387-2419
Cai J and R Walkling 2011 Shareholdersrsquo Say on Pay Does it Create Value Journal of
Financial and Quantitative Analysis 46299ndash339
Coates IV JC (2009) Testimony before the Subcommittee on Securities Insurance and
Investment of the Committee on Banking Housing and Urban Affairs United States Senate
July 29 2009 httppapersssrncomsol3paperscfmabstract_id=1475355
Correa R and U Lel 2013 Say On Pay Laws Executive Compensation CEO Pay Slice and
Firm Value Around the World Federal Reserve Board Working Paper
Cuntildeat V M Gine and M Guadalupe 2012 The Vote is Cast The Effect of Corporate
Governance on Shareholder Value Journal of Finance 671943-1977
Cuntildeat V M Gine and M Guadalupe 2013 Say Pays Shareholder Voice and Firm
Performance ECGI - Finance Working Paper No 373
Del Guercio D L Seery and T Woidtke 2008 Do Boards Pay Attention When Institutional
Investors lsquoJust Vote Norsquo Journal of Financial Economics 9084-103
Ertimur Y F Ferri and S Stubben 2010 Board of Directorslsquo Responsiveness to Shareholders
Evidence from Shareholder Proposals Journal of Corporate Finance 1653-72
Ertimur Y F Ferri and V Muslu 2011 Shareholder Activism and CEO Pay Review of
Financial Studies 24535ndash592
Ertimur Y F Ferri and D Oesch 2013 Shareholder Votes and Proxy Advisors Evidence from
Say on Pay Journal of Accounting Research 51951-996
Ertimur Y F Ferri and D Oesch forthcoming Does the Director Election System Matter
Evidence from Majority Voting Review of Accounting Studies forthcoming
Ferri F 2012 Low-Cost Shareholder Activism A Review of the Evidence Research
Handbook on the Economics of Corporate Law ed CA Hill and BH McDonnell
Edward Elgar Publishing
Ferri F and D Maber 2013 Say on Pay Votes and CEO Compensation Evidence from the UK
Review of Finance 17527-563
Ferri F and D Oesch 2013 Management Influence on Investors Evidence from Shareholder
Votes on the Frequency of Say on Pay Columbia Business School Research Paper No 13-17
Ferri F and T Sandino 2009 The impact of shareholder activism on financial reporting and
compensation The case of employee stock options expensing The Accounting Review 84433ndash
466
Ferri F and J Weber 2009 AFSCME vs Mozilohellipand Say on Pay for All (A) Harvard
Business School Case 109-009
Gompers P J Ishii and A Metrick 2003 Corporate Governance and Equity Prices Quarterly
Journal of Economics 118107-155
Gordon J 2009 Say on Paylsquo Cautionary Notes on the UK Experience and the Case for
Shareholders Opt-In Harvard Journal on Legislation 46323-368
Iliev P and S Vitanova 2013 The effect of the Say on Pay in the US Pennsylvania State
University Working Paper
Kaplan S 2007 Testimony of Steven N Kaplan on Empowering Shareholders on Executive
Compensation and HR 1257 the Shareholder Vote on Executive Compensation Act before the
Committee on Financial Services United States House of Representatives March 8
Larcker D A McCall and G Ormazabal 2013 The Economic Consequences of Proxy
Advisor Say-on-Pay Voting Policies Working Paper Stanford University
Larcker D G Ormazabal and D Taylor 2011 The market reaction to corporate governance
regulation Journal of Financial Economics 101431ndash448
Levit D and N Malenko 2011 Non-binding voting for shareholder proposals Journal of
Finance 661579ndash1614
Mishel L and N Sabadish 2012 How executive compensation and financial-sector pay have
fueled income inequality Issue Brief 331 Economic Policy Institute
Murphy K 2012 The Politics of Pay A Legislative History of Executive Compensation in
Jennifer Hill and Randall Thomas eds Research Handbook on Executive Pay Edward Elgar
Publishers
New York Times 2007 House votes to give investors say on executive pay April 21
Norris F 2004 Corporate democracy and the power to embarrass New York Times March 4
Ozbas O and J Matsusaka2013 Managerial Accommodation Proxy Access and the Cost of
Shareholder Empowerment University of Southern California CLEO Research Paper Series No
C12-1
Thomas R and J Cotter 2007 Shareholder proposals in the new millennium Shareholder
support board response and market reaction Journal of Corporate Finance 13 368ndash391
Thomas R A Palmiter and J Cotter 2012 Dodd-Franks Say on Pay Will it Lead to a Greater
Role for Shareholders in Corporate Governance Cornell Law Review 971213-1266
Thomas R and C Van der Elst 2013 The International Scope of Say on Pay ECGI Law
Working Paper No227
CEO pay to poor performance They also find that this increase does not occur for a subset of
UK firms exempted from the SOP regulation (firms traded on the Alternative Investment Market
a sub-market of the London Stock Exchange with a more flexible regulatory system) suggesting
that the result reflects the impact of SOP rather than a general trend affecting all firms Besides
Ferri and Maber (2013) find that the increase is more pronounced in firms experiencing high
voting dissent and firms with high abnormal CEO pay before the adoption of SOP consistent
with a causal impact of SOP In contrast they fail to find any effect of SOP on the level of pay
Correa and Lel (2013) examine the effect of SOP laws on CEO pay using a large cross-country
sample of about 103000 firm-year observations from 39 countries including 12 countries that
adopted some (advisory or binding) version of SOP They essentially compare post-SOP CEO
pay at firms of countries adopting SOP to a control sample which includes all non-SOP
observations (that is pre-SOP CEO pay in countries eventually adopting SOP as well as CEO
pay in countries never adopting SOP) They find that firms in a post-SOP regime (i) exhibit
lower CEO pay levels (but only in countries adopting an advisory SOP vote) a finding driven by
a relative decline in equity awards and limited to CEOs and not the other top executives (ii)
higher pay-performance sensitivity (the authors do not look at positive and negative performance
separately) The current version of the paper does not examine whether these findings are more
pronounced in (or driven by) firms with excess CEO pay before the adoption of SOP and in
firms experiencing adverse SOP votes
The drawback of the research design in Correa and Lel (2013) is that it is not clear whether the
results reflect time-series changes within countries adopting SOP (pre vs post) or difference
between firms in SOP and non-SOP countries For example further tests in the paper suggest
that the lower CEO pay levels for post-SOP observations are not driven by a pay decrease
relative to the pre-SOP period (in fact country-by-country analyses indicate no change in CEO
pay levels after the adoption of SOP) but by the comparison with non-SOP countries Also while
the authors employ reasonable econometric solutions to deal with this issue some concerns
remains as to whether non-SOP countries can be a good benchmark for what would have
happened in SOP countries had SOP not been adopted (in other words if only countries with
problematic CEO practices adopt SOP it is possible that changes in CEO pay reflect factors
leading to the adoption of SOP rather than the effect of SOP laws per se) Notwithstanding these
concerns these findings are intriguing and future versions of this recent study have the potential
to provide more conclusive evidence on the effect of SOP
While in the US SOP was mandated by the Dodd-Frank Act its implementation was left to the
Securities and Exchange Commission (SEC) The final rule issued in January 2011 exempted
lsquosmallrsquo firms (ie firms with a public float below $75 million) from the SOP provision for two
years Iliev and Vitanova (2013) exploit this setting to estimate the effect of SOP on CEO
compensation In particularly using a regression discontinuity design they compare changes in
CEO pay of firms just above the SEC-imposed threshold to changes in CEO pay of firms just
below the threshold (but otherwise quite similar) They find no differences in terms of level and
composition of CEO pay and in terms of golden parachutes concluding that SOP had no impact
on CEO pay While the identification strategy chosen by the authors has the benefit of reducing
endogeneity issues it comes with the price of focusing only on fairly small firms (those around
the $75 million public float threshold) where CEO pay may not be a problem in the first place
(in which case the lack of impact of SOP would not be surprising but would not speak to its
impact at large firms traditionally the target of compensation-related criticism)
Cuntildeat Gine and Guadalupe (2013) also use a regression discontinuity design but in a different
setting In particular they analyze nonbinding shareholder proposals to adopt SOP (250
proposals between 2006 and 2010) Comparing the changes in CEO pay for firms voluntarily
adopting SOP in response to the proposals to firms not adopting SOP would be difficult since
the adoption decision is endogenous Similarly it would be difficult to compare changes in CEO
pay between firms where the proposal is passed (and thus significantly more likely to be
adopted) and firms where the proposal fails to be approved since the voting outcome is
endogenous as well
Hence Cuntildeat Gine and Guadalupe (2013) use instead a regression discontinuity design that
compares firms where the proposals receive slightly more than 50 of the votes (the threshold
for approval) to otherwise similar firms where the proposals receive slightly less than 50
Using this approach they conclude that the passing and adoption of SOP proposals is not
associated with changes in the level of CEO pay nor its composition and structure (mix of cash
and equity equity incentives level of perks and deferred pay) In some sense this finding
complements the one in Iliev and Vitanova (2013) in that it is based on large firms (shareholder
proposals are typically submitted at large SampP 500 firms) However both Cai and Walkling
(2011) and Cuntildeat Gine and Guadalupe (2013) show that firms targeted by SOP proposals do not
exhibit excess CEO pay or weaker governance8 Hence the lack of an effect at these firms may
8 Activists submitting SOP proposals chose to focus on a broad sample of large firms rather than target only firms
with problematic CEO pay practices to obtain visibility and show to policy-makers that investorrsquos support for SOP
was not limited to problematic firms (Ferri and Weber 2009)
not be surprising More generally the findings cannot be necessarily generalized to firms away
from the 50 threshold or to firms not targeted by SOP shareholder proposals
B Effect of SOP votes on compensation practices
Analyses of the effect of SOP on level and composition of CEO pay may not capture a number
of other changes to compensation contracts that may be induced by SOP votes but are not
reflected in measure of CEO pay For example the introduction of performance-based vesting
provisions is typically not reflected in the estimates of fair value of equity grants used by
researchers in measuring CEO pay Many other changes (eg terms of severance packages) will
only be reflected in measures of CEO pay contingent upon the occurrence of certain events
Hence it is important to complement the evidence in section IIA with an analysis of firmsrsquo
disclosures of changes to observable provisions of compensation contracts Another benefit of
this approach is that it captures the changes that boards explicitly present as a response to SOP
votes and thus it acts as a reality check on regression-based inferences For example a
regression-based finding of say a decrease in CEO pay level would be more credible if
supported by evidence that firms disclose a decision to reduce target levels of CEO pay in
response to an adverse SOP vote (an action that presumably they have an incentive to disclose to
get rewarded by shareholders in subsequent votes)
Two studies collect data on specific changes made to compensation contracts explicitly in
response to SOP votes in the UK and the US using firmsrsquo disclosures in their proxy statements
Ferri and Maber (2013) examine the compensation reports of a sample of UK firms before and
after the first SOP vote and find that firms experiencing higher voting dissent were significantly
more likely to remove compensation provisions criticized by investors as lsquorewards for failurersquo
relative to a matched sample of firms experiencing lower dissent For example they report that
among firms with long notice periods (implying larger severance payments ) the percentage of
high dissent firms that shortened them after the vote (800) was significantly higher than before
the vote (200) and also significantly higher than among low dissent firms after the vote
(333) They report similar findings for another controversial practice namely the presence of
retesting provisions in the performance-based vesting conditions of equity grants Most firms
indicate that these changes were the result of consultations with their major institutional
investors Firmsrsquo responsiveness to SOP votes was lsquorewardedrsquo with a substantial increase in
favorable SOP votes at the subsequent annual meeting Ferri and Maber (2013) also find that
many firms experiencing low voting dissent at the first SOP vote had removed these practices
before the vote9 highlighting the importance of accounting for ex ante effects when assessing the
impact of a new regulation
Ertimur Ferri and Oesch (2013) examine the effect of SOP on compensation practices in the US
in 2011 (first proxy season under mandatory SOP) Their key findings are generally similar to
the evidence from the UK (i) failed SOP votes are rare (about 2 of the sample) though cases
of substantial dissent are more frequent (ii) more than 55 of the firms experiencing significant
voting dissent respond by making material changes to their compensation plan during the
subsequent year (eg introduction of performance-based vesting conditions in equity grants use
of tougher performance targets removal of perks and tax gross-ups removal of controversial
provisions from severance contracts etc) usually in consultation with major institutional
9 In their sample 700 of low dissent firms shortened their notice periods during the year before the SOP vote
investors (iii) firms making changes to their compensation plans receive greater voting support
at the following SOP vote
Ertimur Ferri and Oesch (2013) also highlight the strong influence of the recommendations
released by proxy advisors (particularly Institutional Shareholder Services ISS) with a negative
recommendation being associated with 25 more votes against the compensation plan Further
evidence of the significant influence of ISS is the striking discontinuity in the relation between
the extent of SOP voting dissent and firmsrsquo responsiveness to the vote the percentage of firms
making compensation changes in response to SOP votes jumps from 32 to 72 around the
30 voting dissent threshold Why After the 2011 proxy season ISS had indicated that firms
failing to ldquoadequatelyrdquo respond to SOP voting dissent above 30 would receive a negative
recommendation in 2012 on the SOP proposal and on the election of compensation committee
members
Finally similar to the UK experience there is also anecdotal evidence of analogous
compensation changes being made by US firms ahead of the SOP vote to avoid a negative
proxy advisor recommendation and an adverse shareholder vote (Thomas Palmiter and Cotter
2013 Larcker McCall and Ormazabal 2013)
III The Effect of Say on Pay on Firm Value
In this section we examine the empirical studies on the overall effect of SOP on firm value
A Event studies around the adoption of Say on Pay regulations
As mentioned in the Introduction on April 20 2007 the House of Representatives passed a SOP
Bill by a 2-1 margin On the same day then-Senator Barack Obama introduced a companion
bill (S1181) in the Senate (which was then put on hold by the Senate Banking Committee)
While the SOP Billrsquos approval was expected (Democrats were in control of the House and
supported the SOP Bill) the 2-1 margin was unexpected suggesting some support for SOP
among Republicans
Cai and Walkling (2011) examine the market reaction to the Housersquos approval of the SOP Bill
for a sample of firms in the SampP 1500 index and document a positive reaction for firms more
likely to benefit from greater shareholder voice over executive pay namely firms with high
abnormal CEO cash pay (defined as the difference between actual CEO cash pay and the level
predicted based on known economic determinants such a size performance industry) firms
with low pay-for-performance sensitivity firms with a history of shareholders willing to vote
against compensation-related management proposals and firms with a history of responsiveness
to shareholder pressure on compensation issues In other words they find a positive reaction in
firms where the compensation problems are more severe and a say on pay vote is more likely to
trigger a response (eg more shareholders willing to vote against management and greater firmsrsquo
propensity to respond to an adverse vote) Note however that they do not find a significant
impact for firms with high abnormal equity and total CEO pay
Larcker Ormazabal and Taylor (2011) also examine the market reaction to the Housersquos approval
of the SOP Bill (in addition to other regulatory events related to executive pay and other
governance provisions) Similar to Cai and Wakling (2011) they fail to find a significant impact
for firms with high abnormal total CEO pay (they do not examine the price reaction for firms
with abnormal cash CEO pay or firms with lower pay-performance sensitivity)
A concern with both studies is that the Housersquos approval of a bill does not guarantee passage in
the Senate and approval by the White House In fact the press at the time reported that the
prospects of the bill in the Senate were uncertain (New York Times 2007) and the Bush White
House openly opposed the Bill (Associated Press 2007) Hence it is not clear the extent to
which this event increased the likelihood of a say on pay legislation Arguably as noted in
Section I the only lsquoeventrsquo that substantially increased the likelihood of say on pay legislation
was the financial crisis and the related perception of compensation abuses (eg outrage over AIG
bonuses)
Ferri and Maber (2013) argue that the announcement of the submission of SOP regulation to the
Parliament in the United Kingdom in June 2002 offers a more powerful setting for an event
study since it was largely unexpected and increased substantially the probability of SOP
adoption (Parliamentrsquos approval was virtually guaranteed) Using this event they document
positive abnormal returns for firms with excess CEO pay combined with poor performance and
for firms with controversial pay practices that weaken the penalties for poor performance (those
practices were often removed in response to SOP votes as discussed earlier in Section IIB)
They interpret their findings as consistent with shareholders perceiving SOP as a value
enhancing monitoring mechanism for firms with low pay-to-poor-performance sensitivity
Finally a recent study by Iliev and Vitanova (2013) examines the market reaction around the
announcement of the SEC final SOP rule that on January 25 2011 exempted small firms (ie
firms with a public float below $75 million) from adopting SOP for two years (press reports at
the time noted that the exemption could become permanent) An appealing feature of this setting
is that the SEC decision was an unexpected reversal from the rule proposed in October 2010 rule
under which no publicly traded firm would be exempted
Iliev and Vitanova (2013) compare the market reaction to this announcement for firms just above
the SEC-imposed threshold and thus subject to the SOP rule and firms just below the threshold
(but otherwise quite similar) and thus exempted from SOP While they find not significant
abnormal returns for either group the abnormal returns for exempted firms are significantly
more negative a finding that the authors interpret as evidence of a positive value effect of
mandatory SOP While the setting is an interesting one some concerns remain as exempted
firms are fairly small and the magnitude of the executive pay problem (if any) is unlikely to be
large enough to explain the differential market reaction Also the same study finds no effect on
CEO pay level and composition (see Section IIA) Hence the reason for the differential stock
price reaction remains unclear It would be helpful to know if the event study result is driven by
(or stronger for) exempted firms with excessive pay or problematic pay practices
B Event studies around Shareholder Proposals to Adopt Say on Pay
Starting in 2006 shareholder activists led by union pension funds submitted shareholder
proposal to adopt SOP at hundreds of US firms in an attempt to induce voluntary or mandatory
widespread adoption of SOP
Cai and Walkling (2011) examine the market reaction to proxy filings and annual meetings of
113 firms targeted by nonbinding shareholder proposals to adopt SOP between 2006 and 2008
On average they find insignificant returns around both the submission of the proposal (proxy
filing date) and the vote on the proposal (annual meeting date) But they also find that when the
proposals are filed by union pension funds the abnormal returns (still insignificant) are more
negative than when filed by other activists and that when the proposal is defeated the abnormal
returns (while insignificant) are more positive than when the proposal is passed (it is not clear
whether this result is driven by union-sponsored proposals) They interpret these findings as
evidence that the market views SOP proposals filed by union pension funds are driven by special
interests rather than value maximization10
However caution is required in interpreting this
evidence Virtually all activists filing SOP proposals were coordinated by a group of investors
(mostly but not only union pension funds) who made available to other activists a template for
SOP proposals and a list of target firms (Ferri and Weber 2009) Hence the distinction between
union and non-union proponents in the context of SOP proposals may be only apparent More
importantly the list of target firms was publicly available months before the proxy filing dates
so it is not clear whether the proxy statements contained any new information Similarly the
voting outcome of the SOP proposals may largely be anticipated (based on the composition of
institutional owners proxy advisorsrsquo recommendations etc) Finally the analyses do not control
for other (potentially new) information contained in proxy statements (eg executive
compensation report other items up for a vote at the annual meeting) or other events occurring at
the annual meeting (eg shareholder votes on other items)
Cuntildeat Gine and Guadalupe (2013) also examine the market reaction to the voting outcome of
nonbinding shareholder proposals to adopt SOP but to alleviate these concerns they employ a
lsquofuzzyrsquo regression discontinuity design (RDD) essentially comparing the stock price reaction to
SOP proposals that pass by a small margin to the reaction to SOP proposals that fail by a small
10
Consistent with this interpretation the authors also find that SOP proposals generally targeted larger firms rather
than firms with excessive pay or lower pay-performance sensitivity However as noted by Ferri and Sandino (2009)
activists typically submit shareholder proposals aimed at promoting policy reforms at a broad sample of large firms
rather than the firms that would most benefit from the proposals because they believe that showing widespread
support for the proposal across all types of firms enhances its credibility with policy makers
margin (Cuntildeat Gine and Guadalupe 2012 Ertimur Ferri and Oesch forthcoming) The underlying idea
is that firms around the threshold are likely to have similar characteristics (indeed they do as the
authors show) but differ in the likelihood of implementation which is typically much higher for
proposals passing the threshold (Ertimur Ferri and Stubben 2010 Thomas and Cotter 2007)
Indeed the authors show that the likelihood of subsequently adopting SOP is 48 higher when
the SOP proposal passes by a small margin relatively to when it fails by a small margin Besides
because in these close-call situations the voting outcome is uncertain its resolution (pass or fail)
is likely to convey new information about the likelihood of SOP adoption making this setting
suitable to an event study11
Using this RDD approach Cuntildeat Gine and Guadalupe (2013) find that on the day of the vote a
SOP proposal that passes by a small margin yields an abnormal return of 24 relative to one
that fails (after controlling for other proposals voted upon at the same meeting) and estimate the
lsquofullrsquo value of SOP at 46 (after taking into account the increase in the probability of SOP
implementation)12
Aside from the issues discussed in Section IA (ie generalizability to other
firms) this estimate seems too large to reflect the present value of future reductions in excess
CEO pay (as noted by the authors their estimate of the value of SOP is equivalent to the
estimated value of removing two anti-takeover provision based on Cuntildeat Gine and Guadalupe
2012) particularly because (i) the authors find no evidence of subsequent changes in levels and
composition of CEO pay for firms implementing SOP (see Section IIA) an (ii) the sample firms
11
Consistent with this observation the authors find no market reaction to the voting outcome of SOP shareholder
proposals away from the threshold
12 Since the outcome of the vote is not binding the 24 market reaction only reflects the expected increase in the
probability of SOP adoption and thus understates the value of the SOP provision
do not appear to be characterized by excess CEO pay However the authors also find significant
improvements in subsequent operating performance and efficiency as a result of passing SOP
proposals and thus conjecture that the SOP vote is viewed as a tool to express a vote of
confidence in management performance more than a way to change compensation practices and
that the large positive market reaction reflects expected performance improvements under this
tighter monitoring regime While this is an interesting idea it remains unclear what additional
ldquoteethrdquo a SOP vote may provide over existing (non-compensation related) monitoring
mechanisms shareholders can express (and do often express) their dissatisfaction with
management performance when voting on director elections (Cai Garner and Walkling 2009)
C Event studies around compensation changes induced by SOP
Two studies examine announcements of changes to compensation plans related to SOP Larcker
McCall and Ormazabal (2013) examine the market reaction to (non-contaminated) compensation
changes disclosed in 8-K filings by Russell 3000 firms in the US during the year before the first
SOP vote They find that changes that appear to be made to avoid a negative proxy advisorrsquos
recommendation and thus a negative SOP vote are associated with a negative abnormal return
of -044 whereas other compensation changes are not associated with a significant stock price
reaction Ertimur Ferri and Oesch (2013) perform a similar analysis but focusing on
compensation changes made after the first SOP vote (and explicitly in response to the vote) by
firms receiving a negative recommendation and a high voting dissent in 2011 They fail to find
significant abnormal returns even for the subset of compensation changes that resulted in a
positive recommendation and low dissent in 2012 (and thus were presumably perceived to be
adequate and material by proxy advisors and voting shareholders)
D Other approaches effect of SOP on Tobinrsquos Q
In their cross-country study Correa and Lel (2013) also examine the effect of SOP laws on
Tobinrsquos Q and report a 36 increase in firm value following the adoption of the SOP laws
(more precisely a 36 higher firm value for post-SOP observations relative to a control sample
that pools together pre-SOP and non-SOP observations) They acknowledge that this increase in
firm value is too large to be justified by the relative decrease in CEO pay that they document and
suggest (but do not test) that it may reflect better alignment of pay and performance13
A key
concern is that higher valuation in countries with SOP laws may reflect other governance
changes introduced at the same time While the study controls for other compensation-related
laws there may be other non-compensation regulations introduced with SOP and with a
potentially larger impact on firm value
IV What have we learned
Adoption of SOP in the US has been long advocated by those who contend that executive pay is
a manifestation of rather than a solution to the agency problem and by those who favor great
shareholder involvement in corporate decisions A growing body of research is starting to
investigate the effect of SOP on executive pay and firm value It is premature to draw definitive
conclusions also because these studies differ in methodologies and settings but three points
seem to emerge First there is robust evidence that boards respond to SOP votes when
13
They also find that the firm value increase is higher when the relative decrease in CEO pay results in a lower pay
differential between CEO and other top managers (CEO pay slice) implying that a source of value creation may be
the reduced pay inequality among the top management team But this seems unlikely to explain the change in
Tobinrsquos Q since the increase occurs for both firms with advisory SOP laws and firms with binding SOP laws and
there is no change in CEO pay slice in firms subject to binding SOP laws
shareholders use it ldquovoicerdquo is heardrdquo Both in the US and UK studies show that not only firms
failing to win the SOP vote but also firms facing substantial dissent (20-30 of the votes
against) make changes to their compensation plans in consultation with institutional investors
and proxy advisors Second in both the US and UK institutional investors appear to use the
power of SOP votes to pressure firms into strengthening the link (or the perceived link) between
pay and performance (particularly on the downside) but not to pressure firms to reduce target
levels of pay suggesting a reluctance of institutional investors to lsquoregulatersquo executive pay Third
and consistent with the above most studies examining the aggregate effect of SOP on CEO pay
find some increase in pay-performance sensitivity but not effect on the level and growth of CEO
pay While some observers may interpret the lack of an effect on CEO pay levels as evidence of
the failure of SOP one should note that SOP per se is a neutral tool Its impact depends on what
investors choose to lsquosay on payrsquo Fourth based on my interpretation of the existing research to
date there is little evidence of an effect of SOP on firm value The studies documenting a
positive price impact appear subject to alternative interpretations or not plausible in their
findings (eg they fail to find an effect on CEO pay or the effect is too small to justify the
documented price impact hence it remains unclear what the source of value creation is) Also
there seems to be no stock price reaction to the SOP-induced compensation changes Finally
given the nature of these SOP-induced changes while some of them may be beneficial it seems
hard to believe that they will have a statistically detectable impact on firm value except perhaps
in a handful of firms where the compensation plan is subject to a complete overhaul
Perhaps one explanation for the weak impact of SOP is that executive pay problems were
overstated or that by the time SOP was introduced (more than a decade after the Enron-type
scandals that led to calls for greater shareholder voice) they had been already addressed via
other mechanisms (hedge fund activism monitoring by institutional investors vote-no
campaigns against compensation committee members SEC-mandated pay disclosures)
Overall though based on the evidence to date it does not appear that SOP had a major impact14
Was the adoption of SOP an lsquooptimalrsquo choice from a social welfare point of view This is a
difficult question to answer given the challenges of identifying and measuring all the potential
costs and benefits associated with SOP (likely to change over time and differ across countries)
Perhaps the most positive view of SOP is that its introduction prevented the adoption of other
more radical and intrusive regulatory measures (eg CEO pay caps) In that sense it may be
viewed as an lsquooptimalrsquo answer (ie the lesser of two evils) to the political pressure to reform
executive pay during the financial crisis
14
This interpretation of the evidence is also consistent with the prediction of analytical studies Ozbas and
Matsusaka (2013) model the benefits and costs of shareholder empowerment distinguishing between the right to
approve and the right to propose They show that permitting shareholders to propose directors or policies can cause
value-maximizing managers to take value-reducing actions to accommodate activist investors with non-value-
maximizing goals As for the right to approve it is weakly beneficial that is approval rights (such as a SOP vote)
are generally beneficial for shareholders in that they limit the managerrsquos ability to pursue private benefits at
shareholder expense but they have minimal impact on managerial actions and firm value (because the manager in
effect can threaten shareholders with an undesirable status quo if they do not approve the managerrsquos proposed
action)
References
Associated Press 2007 Administration opposes say on pay bill April 19
Bainbridge S 2008 Remarks on say on pay An unjustified incursion on director authority
UCLA School of Law Research Paper No 08-06
Bebchuk L 2005 The Case for Increasing Shareholder Power Harvard Law Review 118833-
917
Bebchuk L 2007 Written testimony submitted before the Committee on Financial Services
United States House of Representatives Hearing on Empowering Shareholders on Executive
Compensation March 8
Bebchuk LA A Cohen and A Ferrell 2009 What Matters in Corporate Governance Review
of Financial Studies 22783-827
Cai J J Garner and R Walkling 2009 Electing Directors Journal of Finance 642387-2419
Cai J and R Walkling 2011 Shareholdersrsquo Say on Pay Does it Create Value Journal of
Financial and Quantitative Analysis 46299ndash339
Coates IV JC (2009) Testimony before the Subcommittee on Securities Insurance and
Investment of the Committee on Banking Housing and Urban Affairs United States Senate
July 29 2009 httppapersssrncomsol3paperscfmabstract_id=1475355
Correa R and U Lel 2013 Say On Pay Laws Executive Compensation CEO Pay Slice and
Firm Value Around the World Federal Reserve Board Working Paper
Cuntildeat V M Gine and M Guadalupe 2012 The Vote is Cast The Effect of Corporate
Governance on Shareholder Value Journal of Finance 671943-1977
Cuntildeat V M Gine and M Guadalupe 2013 Say Pays Shareholder Voice and Firm
Performance ECGI - Finance Working Paper No 373
Del Guercio D L Seery and T Woidtke 2008 Do Boards Pay Attention When Institutional
Investors lsquoJust Vote Norsquo Journal of Financial Economics 9084-103
Ertimur Y F Ferri and S Stubben 2010 Board of Directorslsquo Responsiveness to Shareholders
Evidence from Shareholder Proposals Journal of Corporate Finance 1653-72
Ertimur Y F Ferri and V Muslu 2011 Shareholder Activism and CEO Pay Review of
Financial Studies 24535ndash592
Ertimur Y F Ferri and D Oesch 2013 Shareholder Votes and Proxy Advisors Evidence from
Say on Pay Journal of Accounting Research 51951-996
Ertimur Y F Ferri and D Oesch forthcoming Does the Director Election System Matter
Evidence from Majority Voting Review of Accounting Studies forthcoming
Ferri F 2012 Low-Cost Shareholder Activism A Review of the Evidence Research
Handbook on the Economics of Corporate Law ed CA Hill and BH McDonnell
Edward Elgar Publishing
Ferri F and D Maber 2013 Say on Pay Votes and CEO Compensation Evidence from the UK
Review of Finance 17527-563
Ferri F and D Oesch 2013 Management Influence on Investors Evidence from Shareholder
Votes on the Frequency of Say on Pay Columbia Business School Research Paper No 13-17
Ferri F and T Sandino 2009 The impact of shareholder activism on financial reporting and
compensation The case of employee stock options expensing The Accounting Review 84433ndash
466
Ferri F and J Weber 2009 AFSCME vs Mozilohellipand Say on Pay for All (A) Harvard
Business School Case 109-009
Gompers P J Ishii and A Metrick 2003 Corporate Governance and Equity Prices Quarterly
Journal of Economics 118107-155
Gordon J 2009 Say on Paylsquo Cautionary Notes on the UK Experience and the Case for
Shareholders Opt-In Harvard Journal on Legislation 46323-368
Iliev P and S Vitanova 2013 The effect of the Say on Pay in the US Pennsylvania State
University Working Paper
Kaplan S 2007 Testimony of Steven N Kaplan on Empowering Shareholders on Executive
Compensation and HR 1257 the Shareholder Vote on Executive Compensation Act before the
Committee on Financial Services United States House of Representatives March 8
Larcker D A McCall and G Ormazabal 2013 The Economic Consequences of Proxy
Advisor Say-on-Pay Voting Policies Working Paper Stanford University
Larcker D G Ormazabal and D Taylor 2011 The market reaction to corporate governance
regulation Journal of Financial Economics 101431ndash448
Levit D and N Malenko 2011 Non-binding voting for shareholder proposals Journal of
Finance 661579ndash1614
Mishel L and N Sabadish 2012 How executive compensation and financial-sector pay have
fueled income inequality Issue Brief 331 Economic Policy Institute
Murphy K 2012 The Politics of Pay A Legislative History of Executive Compensation in
Jennifer Hill and Randall Thomas eds Research Handbook on Executive Pay Edward Elgar
Publishers
New York Times 2007 House votes to give investors say on executive pay April 21
Norris F 2004 Corporate democracy and the power to embarrass New York Times March 4
Ozbas O and J Matsusaka2013 Managerial Accommodation Proxy Access and the Cost of
Shareholder Empowerment University of Southern California CLEO Research Paper Series No
C12-1
Thomas R and J Cotter 2007 Shareholder proposals in the new millennium Shareholder
support board response and market reaction Journal of Corporate Finance 13 368ndash391
Thomas R A Palmiter and J Cotter 2012 Dodd-Franks Say on Pay Will it Lead to a Greater
Role for Shareholders in Corporate Governance Cornell Law Review 971213-1266
Thomas R and C Van der Elst 2013 The International Scope of Say on Pay ECGI Law
Working Paper No227
that the lower CEO pay levels for post-SOP observations are not driven by a pay decrease
relative to the pre-SOP period (in fact country-by-country analyses indicate no change in CEO
pay levels after the adoption of SOP) but by the comparison with non-SOP countries Also while
the authors employ reasonable econometric solutions to deal with this issue some concerns
remains as to whether non-SOP countries can be a good benchmark for what would have
happened in SOP countries had SOP not been adopted (in other words if only countries with
problematic CEO practices adopt SOP it is possible that changes in CEO pay reflect factors
leading to the adoption of SOP rather than the effect of SOP laws per se) Notwithstanding these
concerns these findings are intriguing and future versions of this recent study have the potential
to provide more conclusive evidence on the effect of SOP
While in the US SOP was mandated by the Dodd-Frank Act its implementation was left to the
Securities and Exchange Commission (SEC) The final rule issued in January 2011 exempted
lsquosmallrsquo firms (ie firms with a public float below $75 million) from the SOP provision for two
years Iliev and Vitanova (2013) exploit this setting to estimate the effect of SOP on CEO
compensation In particularly using a regression discontinuity design they compare changes in
CEO pay of firms just above the SEC-imposed threshold to changes in CEO pay of firms just
below the threshold (but otherwise quite similar) They find no differences in terms of level and
composition of CEO pay and in terms of golden parachutes concluding that SOP had no impact
on CEO pay While the identification strategy chosen by the authors has the benefit of reducing
endogeneity issues it comes with the price of focusing only on fairly small firms (those around
the $75 million public float threshold) where CEO pay may not be a problem in the first place
(in which case the lack of impact of SOP would not be surprising but would not speak to its
impact at large firms traditionally the target of compensation-related criticism)
Cuntildeat Gine and Guadalupe (2013) also use a regression discontinuity design but in a different
setting In particular they analyze nonbinding shareholder proposals to adopt SOP (250
proposals between 2006 and 2010) Comparing the changes in CEO pay for firms voluntarily
adopting SOP in response to the proposals to firms not adopting SOP would be difficult since
the adoption decision is endogenous Similarly it would be difficult to compare changes in CEO
pay between firms where the proposal is passed (and thus significantly more likely to be
adopted) and firms where the proposal fails to be approved since the voting outcome is
endogenous as well
Hence Cuntildeat Gine and Guadalupe (2013) use instead a regression discontinuity design that
compares firms where the proposals receive slightly more than 50 of the votes (the threshold
for approval) to otherwise similar firms where the proposals receive slightly less than 50
Using this approach they conclude that the passing and adoption of SOP proposals is not
associated with changes in the level of CEO pay nor its composition and structure (mix of cash
and equity equity incentives level of perks and deferred pay) In some sense this finding
complements the one in Iliev and Vitanova (2013) in that it is based on large firms (shareholder
proposals are typically submitted at large SampP 500 firms) However both Cai and Walkling
(2011) and Cuntildeat Gine and Guadalupe (2013) show that firms targeted by SOP proposals do not
exhibit excess CEO pay or weaker governance8 Hence the lack of an effect at these firms may
8 Activists submitting SOP proposals chose to focus on a broad sample of large firms rather than target only firms
with problematic CEO pay practices to obtain visibility and show to policy-makers that investorrsquos support for SOP
was not limited to problematic firms (Ferri and Weber 2009)
not be surprising More generally the findings cannot be necessarily generalized to firms away
from the 50 threshold or to firms not targeted by SOP shareholder proposals
B Effect of SOP votes on compensation practices
Analyses of the effect of SOP on level and composition of CEO pay may not capture a number
of other changes to compensation contracts that may be induced by SOP votes but are not
reflected in measure of CEO pay For example the introduction of performance-based vesting
provisions is typically not reflected in the estimates of fair value of equity grants used by
researchers in measuring CEO pay Many other changes (eg terms of severance packages) will
only be reflected in measures of CEO pay contingent upon the occurrence of certain events
Hence it is important to complement the evidence in section IIA with an analysis of firmsrsquo
disclosures of changes to observable provisions of compensation contracts Another benefit of
this approach is that it captures the changes that boards explicitly present as a response to SOP
votes and thus it acts as a reality check on regression-based inferences For example a
regression-based finding of say a decrease in CEO pay level would be more credible if
supported by evidence that firms disclose a decision to reduce target levels of CEO pay in
response to an adverse SOP vote (an action that presumably they have an incentive to disclose to
get rewarded by shareholders in subsequent votes)
Two studies collect data on specific changes made to compensation contracts explicitly in
response to SOP votes in the UK and the US using firmsrsquo disclosures in their proxy statements
Ferri and Maber (2013) examine the compensation reports of a sample of UK firms before and
after the first SOP vote and find that firms experiencing higher voting dissent were significantly
more likely to remove compensation provisions criticized by investors as lsquorewards for failurersquo
relative to a matched sample of firms experiencing lower dissent For example they report that
among firms with long notice periods (implying larger severance payments ) the percentage of
high dissent firms that shortened them after the vote (800) was significantly higher than before
the vote (200) and also significantly higher than among low dissent firms after the vote
(333) They report similar findings for another controversial practice namely the presence of
retesting provisions in the performance-based vesting conditions of equity grants Most firms
indicate that these changes were the result of consultations with their major institutional
investors Firmsrsquo responsiveness to SOP votes was lsquorewardedrsquo with a substantial increase in
favorable SOP votes at the subsequent annual meeting Ferri and Maber (2013) also find that
many firms experiencing low voting dissent at the first SOP vote had removed these practices
before the vote9 highlighting the importance of accounting for ex ante effects when assessing the
impact of a new regulation
Ertimur Ferri and Oesch (2013) examine the effect of SOP on compensation practices in the US
in 2011 (first proxy season under mandatory SOP) Their key findings are generally similar to
the evidence from the UK (i) failed SOP votes are rare (about 2 of the sample) though cases
of substantial dissent are more frequent (ii) more than 55 of the firms experiencing significant
voting dissent respond by making material changes to their compensation plan during the
subsequent year (eg introduction of performance-based vesting conditions in equity grants use
of tougher performance targets removal of perks and tax gross-ups removal of controversial
provisions from severance contracts etc) usually in consultation with major institutional
9 In their sample 700 of low dissent firms shortened their notice periods during the year before the SOP vote
investors (iii) firms making changes to their compensation plans receive greater voting support
at the following SOP vote
Ertimur Ferri and Oesch (2013) also highlight the strong influence of the recommendations
released by proxy advisors (particularly Institutional Shareholder Services ISS) with a negative
recommendation being associated with 25 more votes against the compensation plan Further
evidence of the significant influence of ISS is the striking discontinuity in the relation between
the extent of SOP voting dissent and firmsrsquo responsiveness to the vote the percentage of firms
making compensation changes in response to SOP votes jumps from 32 to 72 around the
30 voting dissent threshold Why After the 2011 proxy season ISS had indicated that firms
failing to ldquoadequatelyrdquo respond to SOP voting dissent above 30 would receive a negative
recommendation in 2012 on the SOP proposal and on the election of compensation committee
members
Finally similar to the UK experience there is also anecdotal evidence of analogous
compensation changes being made by US firms ahead of the SOP vote to avoid a negative
proxy advisor recommendation and an adverse shareholder vote (Thomas Palmiter and Cotter
2013 Larcker McCall and Ormazabal 2013)
III The Effect of Say on Pay on Firm Value
In this section we examine the empirical studies on the overall effect of SOP on firm value
A Event studies around the adoption of Say on Pay regulations
As mentioned in the Introduction on April 20 2007 the House of Representatives passed a SOP
Bill by a 2-1 margin On the same day then-Senator Barack Obama introduced a companion
bill (S1181) in the Senate (which was then put on hold by the Senate Banking Committee)
While the SOP Billrsquos approval was expected (Democrats were in control of the House and
supported the SOP Bill) the 2-1 margin was unexpected suggesting some support for SOP
among Republicans
Cai and Walkling (2011) examine the market reaction to the Housersquos approval of the SOP Bill
for a sample of firms in the SampP 1500 index and document a positive reaction for firms more
likely to benefit from greater shareholder voice over executive pay namely firms with high
abnormal CEO cash pay (defined as the difference between actual CEO cash pay and the level
predicted based on known economic determinants such a size performance industry) firms
with low pay-for-performance sensitivity firms with a history of shareholders willing to vote
against compensation-related management proposals and firms with a history of responsiveness
to shareholder pressure on compensation issues In other words they find a positive reaction in
firms where the compensation problems are more severe and a say on pay vote is more likely to
trigger a response (eg more shareholders willing to vote against management and greater firmsrsquo
propensity to respond to an adverse vote) Note however that they do not find a significant
impact for firms with high abnormal equity and total CEO pay
Larcker Ormazabal and Taylor (2011) also examine the market reaction to the Housersquos approval
of the SOP Bill (in addition to other regulatory events related to executive pay and other
governance provisions) Similar to Cai and Wakling (2011) they fail to find a significant impact
for firms with high abnormal total CEO pay (they do not examine the price reaction for firms
with abnormal cash CEO pay or firms with lower pay-performance sensitivity)
A concern with both studies is that the Housersquos approval of a bill does not guarantee passage in
the Senate and approval by the White House In fact the press at the time reported that the
prospects of the bill in the Senate were uncertain (New York Times 2007) and the Bush White
House openly opposed the Bill (Associated Press 2007) Hence it is not clear the extent to
which this event increased the likelihood of a say on pay legislation Arguably as noted in
Section I the only lsquoeventrsquo that substantially increased the likelihood of say on pay legislation
was the financial crisis and the related perception of compensation abuses (eg outrage over AIG
bonuses)
Ferri and Maber (2013) argue that the announcement of the submission of SOP regulation to the
Parliament in the United Kingdom in June 2002 offers a more powerful setting for an event
study since it was largely unexpected and increased substantially the probability of SOP
adoption (Parliamentrsquos approval was virtually guaranteed) Using this event they document
positive abnormal returns for firms with excess CEO pay combined with poor performance and
for firms with controversial pay practices that weaken the penalties for poor performance (those
practices were often removed in response to SOP votes as discussed earlier in Section IIB)
They interpret their findings as consistent with shareholders perceiving SOP as a value
enhancing monitoring mechanism for firms with low pay-to-poor-performance sensitivity
Finally a recent study by Iliev and Vitanova (2013) examines the market reaction around the
announcement of the SEC final SOP rule that on January 25 2011 exempted small firms (ie
firms with a public float below $75 million) from adopting SOP for two years (press reports at
the time noted that the exemption could become permanent) An appealing feature of this setting
is that the SEC decision was an unexpected reversal from the rule proposed in October 2010 rule
under which no publicly traded firm would be exempted
Iliev and Vitanova (2013) compare the market reaction to this announcement for firms just above
the SEC-imposed threshold and thus subject to the SOP rule and firms just below the threshold
(but otherwise quite similar) and thus exempted from SOP While they find not significant
abnormal returns for either group the abnormal returns for exempted firms are significantly
more negative a finding that the authors interpret as evidence of a positive value effect of
mandatory SOP While the setting is an interesting one some concerns remain as exempted
firms are fairly small and the magnitude of the executive pay problem (if any) is unlikely to be
large enough to explain the differential market reaction Also the same study finds no effect on
CEO pay level and composition (see Section IIA) Hence the reason for the differential stock
price reaction remains unclear It would be helpful to know if the event study result is driven by
(or stronger for) exempted firms with excessive pay or problematic pay practices
B Event studies around Shareholder Proposals to Adopt Say on Pay
Starting in 2006 shareholder activists led by union pension funds submitted shareholder
proposal to adopt SOP at hundreds of US firms in an attempt to induce voluntary or mandatory
widespread adoption of SOP
Cai and Walkling (2011) examine the market reaction to proxy filings and annual meetings of
113 firms targeted by nonbinding shareholder proposals to adopt SOP between 2006 and 2008
On average they find insignificant returns around both the submission of the proposal (proxy
filing date) and the vote on the proposal (annual meeting date) But they also find that when the
proposals are filed by union pension funds the abnormal returns (still insignificant) are more
negative than when filed by other activists and that when the proposal is defeated the abnormal
returns (while insignificant) are more positive than when the proposal is passed (it is not clear
whether this result is driven by union-sponsored proposals) They interpret these findings as
evidence that the market views SOP proposals filed by union pension funds are driven by special
interests rather than value maximization10
However caution is required in interpreting this
evidence Virtually all activists filing SOP proposals were coordinated by a group of investors
(mostly but not only union pension funds) who made available to other activists a template for
SOP proposals and a list of target firms (Ferri and Weber 2009) Hence the distinction between
union and non-union proponents in the context of SOP proposals may be only apparent More
importantly the list of target firms was publicly available months before the proxy filing dates
so it is not clear whether the proxy statements contained any new information Similarly the
voting outcome of the SOP proposals may largely be anticipated (based on the composition of
institutional owners proxy advisorsrsquo recommendations etc) Finally the analyses do not control
for other (potentially new) information contained in proxy statements (eg executive
compensation report other items up for a vote at the annual meeting) or other events occurring at
the annual meeting (eg shareholder votes on other items)
Cuntildeat Gine and Guadalupe (2013) also examine the market reaction to the voting outcome of
nonbinding shareholder proposals to adopt SOP but to alleviate these concerns they employ a
lsquofuzzyrsquo regression discontinuity design (RDD) essentially comparing the stock price reaction to
SOP proposals that pass by a small margin to the reaction to SOP proposals that fail by a small
10
Consistent with this interpretation the authors also find that SOP proposals generally targeted larger firms rather
than firms with excessive pay or lower pay-performance sensitivity However as noted by Ferri and Sandino (2009)
activists typically submit shareholder proposals aimed at promoting policy reforms at a broad sample of large firms
rather than the firms that would most benefit from the proposals because they believe that showing widespread
support for the proposal across all types of firms enhances its credibility with policy makers
margin (Cuntildeat Gine and Guadalupe 2012 Ertimur Ferri and Oesch forthcoming) The underlying idea
is that firms around the threshold are likely to have similar characteristics (indeed they do as the
authors show) but differ in the likelihood of implementation which is typically much higher for
proposals passing the threshold (Ertimur Ferri and Stubben 2010 Thomas and Cotter 2007)
Indeed the authors show that the likelihood of subsequently adopting SOP is 48 higher when
the SOP proposal passes by a small margin relatively to when it fails by a small margin Besides
because in these close-call situations the voting outcome is uncertain its resolution (pass or fail)
is likely to convey new information about the likelihood of SOP adoption making this setting
suitable to an event study11
Using this RDD approach Cuntildeat Gine and Guadalupe (2013) find that on the day of the vote a
SOP proposal that passes by a small margin yields an abnormal return of 24 relative to one
that fails (after controlling for other proposals voted upon at the same meeting) and estimate the
lsquofullrsquo value of SOP at 46 (after taking into account the increase in the probability of SOP
implementation)12
Aside from the issues discussed in Section IA (ie generalizability to other
firms) this estimate seems too large to reflect the present value of future reductions in excess
CEO pay (as noted by the authors their estimate of the value of SOP is equivalent to the
estimated value of removing two anti-takeover provision based on Cuntildeat Gine and Guadalupe
2012) particularly because (i) the authors find no evidence of subsequent changes in levels and
composition of CEO pay for firms implementing SOP (see Section IIA) an (ii) the sample firms
11
Consistent with this observation the authors find no market reaction to the voting outcome of SOP shareholder
proposals away from the threshold
12 Since the outcome of the vote is not binding the 24 market reaction only reflects the expected increase in the
probability of SOP adoption and thus understates the value of the SOP provision
do not appear to be characterized by excess CEO pay However the authors also find significant
improvements in subsequent operating performance and efficiency as a result of passing SOP
proposals and thus conjecture that the SOP vote is viewed as a tool to express a vote of
confidence in management performance more than a way to change compensation practices and
that the large positive market reaction reflects expected performance improvements under this
tighter monitoring regime While this is an interesting idea it remains unclear what additional
ldquoteethrdquo a SOP vote may provide over existing (non-compensation related) monitoring
mechanisms shareholders can express (and do often express) their dissatisfaction with
management performance when voting on director elections (Cai Garner and Walkling 2009)
C Event studies around compensation changes induced by SOP
Two studies examine announcements of changes to compensation plans related to SOP Larcker
McCall and Ormazabal (2013) examine the market reaction to (non-contaminated) compensation
changes disclosed in 8-K filings by Russell 3000 firms in the US during the year before the first
SOP vote They find that changes that appear to be made to avoid a negative proxy advisorrsquos
recommendation and thus a negative SOP vote are associated with a negative abnormal return
of -044 whereas other compensation changes are not associated with a significant stock price
reaction Ertimur Ferri and Oesch (2013) perform a similar analysis but focusing on
compensation changes made after the first SOP vote (and explicitly in response to the vote) by
firms receiving a negative recommendation and a high voting dissent in 2011 They fail to find
significant abnormal returns even for the subset of compensation changes that resulted in a
positive recommendation and low dissent in 2012 (and thus were presumably perceived to be
adequate and material by proxy advisors and voting shareholders)
D Other approaches effect of SOP on Tobinrsquos Q
In their cross-country study Correa and Lel (2013) also examine the effect of SOP laws on
Tobinrsquos Q and report a 36 increase in firm value following the adoption of the SOP laws
(more precisely a 36 higher firm value for post-SOP observations relative to a control sample
that pools together pre-SOP and non-SOP observations) They acknowledge that this increase in
firm value is too large to be justified by the relative decrease in CEO pay that they document and
suggest (but do not test) that it may reflect better alignment of pay and performance13
A key
concern is that higher valuation in countries with SOP laws may reflect other governance
changes introduced at the same time While the study controls for other compensation-related
laws there may be other non-compensation regulations introduced with SOP and with a
potentially larger impact on firm value
IV What have we learned
Adoption of SOP in the US has been long advocated by those who contend that executive pay is
a manifestation of rather than a solution to the agency problem and by those who favor great
shareholder involvement in corporate decisions A growing body of research is starting to
investigate the effect of SOP on executive pay and firm value It is premature to draw definitive
conclusions also because these studies differ in methodologies and settings but three points
seem to emerge First there is robust evidence that boards respond to SOP votes when
13
They also find that the firm value increase is higher when the relative decrease in CEO pay results in a lower pay
differential between CEO and other top managers (CEO pay slice) implying that a source of value creation may be
the reduced pay inequality among the top management team But this seems unlikely to explain the change in
Tobinrsquos Q since the increase occurs for both firms with advisory SOP laws and firms with binding SOP laws and
there is no change in CEO pay slice in firms subject to binding SOP laws
shareholders use it ldquovoicerdquo is heardrdquo Both in the US and UK studies show that not only firms
failing to win the SOP vote but also firms facing substantial dissent (20-30 of the votes
against) make changes to their compensation plans in consultation with institutional investors
and proxy advisors Second in both the US and UK institutional investors appear to use the
power of SOP votes to pressure firms into strengthening the link (or the perceived link) between
pay and performance (particularly on the downside) but not to pressure firms to reduce target
levels of pay suggesting a reluctance of institutional investors to lsquoregulatersquo executive pay Third
and consistent with the above most studies examining the aggregate effect of SOP on CEO pay
find some increase in pay-performance sensitivity but not effect on the level and growth of CEO
pay While some observers may interpret the lack of an effect on CEO pay levels as evidence of
the failure of SOP one should note that SOP per se is a neutral tool Its impact depends on what
investors choose to lsquosay on payrsquo Fourth based on my interpretation of the existing research to
date there is little evidence of an effect of SOP on firm value The studies documenting a
positive price impact appear subject to alternative interpretations or not plausible in their
findings (eg they fail to find an effect on CEO pay or the effect is too small to justify the
documented price impact hence it remains unclear what the source of value creation is) Also
there seems to be no stock price reaction to the SOP-induced compensation changes Finally
given the nature of these SOP-induced changes while some of them may be beneficial it seems
hard to believe that they will have a statistically detectable impact on firm value except perhaps
in a handful of firms where the compensation plan is subject to a complete overhaul
Perhaps one explanation for the weak impact of SOP is that executive pay problems were
overstated or that by the time SOP was introduced (more than a decade after the Enron-type
scandals that led to calls for greater shareholder voice) they had been already addressed via
other mechanisms (hedge fund activism monitoring by institutional investors vote-no
campaigns against compensation committee members SEC-mandated pay disclosures)
Overall though based on the evidence to date it does not appear that SOP had a major impact14
Was the adoption of SOP an lsquooptimalrsquo choice from a social welfare point of view This is a
difficult question to answer given the challenges of identifying and measuring all the potential
costs and benefits associated with SOP (likely to change over time and differ across countries)
Perhaps the most positive view of SOP is that its introduction prevented the adoption of other
more radical and intrusive regulatory measures (eg CEO pay caps) In that sense it may be
viewed as an lsquooptimalrsquo answer (ie the lesser of two evils) to the political pressure to reform
executive pay during the financial crisis
14
This interpretation of the evidence is also consistent with the prediction of analytical studies Ozbas and
Matsusaka (2013) model the benefits and costs of shareholder empowerment distinguishing between the right to
approve and the right to propose They show that permitting shareholders to propose directors or policies can cause
value-maximizing managers to take value-reducing actions to accommodate activist investors with non-value-
maximizing goals As for the right to approve it is weakly beneficial that is approval rights (such as a SOP vote)
are generally beneficial for shareholders in that they limit the managerrsquos ability to pursue private benefits at
shareholder expense but they have minimal impact on managerial actions and firm value (because the manager in
effect can threaten shareholders with an undesirable status quo if they do not approve the managerrsquos proposed
action)
References
Associated Press 2007 Administration opposes say on pay bill April 19
Bainbridge S 2008 Remarks on say on pay An unjustified incursion on director authority
UCLA School of Law Research Paper No 08-06
Bebchuk L 2005 The Case for Increasing Shareholder Power Harvard Law Review 118833-
917
Bebchuk L 2007 Written testimony submitted before the Committee on Financial Services
United States House of Representatives Hearing on Empowering Shareholders on Executive
Compensation March 8
Bebchuk LA A Cohen and A Ferrell 2009 What Matters in Corporate Governance Review
of Financial Studies 22783-827
Cai J J Garner and R Walkling 2009 Electing Directors Journal of Finance 642387-2419
Cai J and R Walkling 2011 Shareholdersrsquo Say on Pay Does it Create Value Journal of
Financial and Quantitative Analysis 46299ndash339
Coates IV JC (2009) Testimony before the Subcommittee on Securities Insurance and
Investment of the Committee on Banking Housing and Urban Affairs United States Senate
July 29 2009 httppapersssrncomsol3paperscfmabstract_id=1475355
Correa R and U Lel 2013 Say On Pay Laws Executive Compensation CEO Pay Slice and
Firm Value Around the World Federal Reserve Board Working Paper
Cuntildeat V M Gine and M Guadalupe 2012 The Vote is Cast The Effect of Corporate
Governance on Shareholder Value Journal of Finance 671943-1977
Cuntildeat V M Gine and M Guadalupe 2013 Say Pays Shareholder Voice and Firm
Performance ECGI - Finance Working Paper No 373
Del Guercio D L Seery and T Woidtke 2008 Do Boards Pay Attention When Institutional
Investors lsquoJust Vote Norsquo Journal of Financial Economics 9084-103
Ertimur Y F Ferri and S Stubben 2010 Board of Directorslsquo Responsiveness to Shareholders
Evidence from Shareholder Proposals Journal of Corporate Finance 1653-72
Ertimur Y F Ferri and V Muslu 2011 Shareholder Activism and CEO Pay Review of
Financial Studies 24535ndash592
Ertimur Y F Ferri and D Oesch 2013 Shareholder Votes and Proxy Advisors Evidence from
Say on Pay Journal of Accounting Research 51951-996
Ertimur Y F Ferri and D Oesch forthcoming Does the Director Election System Matter
Evidence from Majority Voting Review of Accounting Studies forthcoming
Ferri F 2012 Low-Cost Shareholder Activism A Review of the Evidence Research
Handbook on the Economics of Corporate Law ed CA Hill and BH McDonnell
Edward Elgar Publishing
Ferri F and D Maber 2013 Say on Pay Votes and CEO Compensation Evidence from the UK
Review of Finance 17527-563
Ferri F and D Oesch 2013 Management Influence on Investors Evidence from Shareholder
Votes on the Frequency of Say on Pay Columbia Business School Research Paper No 13-17
Ferri F and T Sandino 2009 The impact of shareholder activism on financial reporting and
compensation The case of employee stock options expensing The Accounting Review 84433ndash
466
Ferri F and J Weber 2009 AFSCME vs Mozilohellipand Say on Pay for All (A) Harvard
Business School Case 109-009
Gompers P J Ishii and A Metrick 2003 Corporate Governance and Equity Prices Quarterly
Journal of Economics 118107-155
Gordon J 2009 Say on Paylsquo Cautionary Notes on the UK Experience and the Case for
Shareholders Opt-In Harvard Journal on Legislation 46323-368
Iliev P and S Vitanova 2013 The effect of the Say on Pay in the US Pennsylvania State
University Working Paper
Kaplan S 2007 Testimony of Steven N Kaplan on Empowering Shareholders on Executive
Compensation and HR 1257 the Shareholder Vote on Executive Compensation Act before the
Committee on Financial Services United States House of Representatives March 8
Larcker D A McCall and G Ormazabal 2013 The Economic Consequences of Proxy
Advisor Say-on-Pay Voting Policies Working Paper Stanford University
Larcker D G Ormazabal and D Taylor 2011 The market reaction to corporate governance
regulation Journal of Financial Economics 101431ndash448
Levit D and N Malenko 2011 Non-binding voting for shareholder proposals Journal of
Finance 661579ndash1614
Mishel L and N Sabadish 2012 How executive compensation and financial-sector pay have
fueled income inequality Issue Brief 331 Economic Policy Institute
Murphy K 2012 The Politics of Pay A Legislative History of Executive Compensation in
Jennifer Hill and Randall Thomas eds Research Handbook on Executive Pay Edward Elgar
Publishers
New York Times 2007 House votes to give investors say on executive pay April 21
Norris F 2004 Corporate democracy and the power to embarrass New York Times March 4
Ozbas O and J Matsusaka2013 Managerial Accommodation Proxy Access and the Cost of
Shareholder Empowerment University of Southern California CLEO Research Paper Series No
C12-1
Thomas R and J Cotter 2007 Shareholder proposals in the new millennium Shareholder
support board response and market reaction Journal of Corporate Finance 13 368ndash391
Thomas R A Palmiter and J Cotter 2012 Dodd-Franks Say on Pay Will it Lead to a Greater
Role for Shareholders in Corporate Governance Cornell Law Review 971213-1266
Thomas R and C Van der Elst 2013 The International Scope of Say on Pay ECGI Law
Working Paper No227
(in which case the lack of impact of SOP would not be surprising but would not speak to its
impact at large firms traditionally the target of compensation-related criticism)
Cuntildeat Gine and Guadalupe (2013) also use a regression discontinuity design but in a different
setting In particular they analyze nonbinding shareholder proposals to adopt SOP (250
proposals between 2006 and 2010) Comparing the changes in CEO pay for firms voluntarily
adopting SOP in response to the proposals to firms not adopting SOP would be difficult since
the adoption decision is endogenous Similarly it would be difficult to compare changes in CEO
pay between firms where the proposal is passed (and thus significantly more likely to be
adopted) and firms where the proposal fails to be approved since the voting outcome is
endogenous as well
Hence Cuntildeat Gine and Guadalupe (2013) use instead a regression discontinuity design that
compares firms where the proposals receive slightly more than 50 of the votes (the threshold
for approval) to otherwise similar firms where the proposals receive slightly less than 50
Using this approach they conclude that the passing and adoption of SOP proposals is not
associated with changes in the level of CEO pay nor its composition and structure (mix of cash
and equity equity incentives level of perks and deferred pay) In some sense this finding
complements the one in Iliev and Vitanova (2013) in that it is based on large firms (shareholder
proposals are typically submitted at large SampP 500 firms) However both Cai and Walkling
(2011) and Cuntildeat Gine and Guadalupe (2013) show that firms targeted by SOP proposals do not
exhibit excess CEO pay or weaker governance8 Hence the lack of an effect at these firms may
8 Activists submitting SOP proposals chose to focus on a broad sample of large firms rather than target only firms
with problematic CEO pay practices to obtain visibility and show to policy-makers that investorrsquos support for SOP
was not limited to problematic firms (Ferri and Weber 2009)
not be surprising More generally the findings cannot be necessarily generalized to firms away
from the 50 threshold or to firms not targeted by SOP shareholder proposals
B Effect of SOP votes on compensation practices
Analyses of the effect of SOP on level and composition of CEO pay may not capture a number
of other changes to compensation contracts that may be induced by SOP votes but are not
reflected in measure of CEO pay For example the introduction of performance-based vesting
provisions is typically not reflected in the estimates of fair value of equity grants used by
researchers in measuring CEO pay Many other changes (eg terms of severance packages) will
only be reflected in measures of CEO pay contingent upon the occurrence of certain events
Hence it is important to complement the evidence in section IIA with an analysis of firmsrsquo
disclosures of changes to observable provisions of compensation contracts Another benefit of
this approach is that it captures the changes that boards explicitly present as a response to SOP
votes and thus it acts as a reality check on regression-based inferences For example a
regression-based finding of say a decrease in CEO pay level would be more credible if
supported by evidence that firms disclose a decision to reduce target levels of CEO pay in
response to an adverse SOP vote (an action that presumably they have an incentive to disclose to
get rewarded by shareholders in subsequent votes)
Two studies collect data on specific changes made to compensation contracts explicitly in
response to SOP votes in the UK and the US using firmsrsquo disclosures in their proxy statements
Ferri and Maber (2013) examine the compensation reports of a sample of UK firms before and
after the first SOP vote and find that firms experiencing higher voting dissent were significantly
more likely to remove compensation provisions criticized by investors as lsquorewards for failurersquo
relative to a matched sample of firms experiencing lower dissent For example they report that
among firms with long notice periods (implying larger severance payments ) the percentage of
high dissent firms that shortened them after the vote (800) was significantly higher than before
the vote (200) and also significantly higher than among low dissent firms after the vote
(333) They report similar findings for another controversial practice namely the presence of
retesting provisions in the performance-based vesting conditions of equity grants Most firms
indicate that these changes were the result of consultations with their major institutional
investors Firmsrsquo responsiveness to SOP votes was lsquorewardedrsquo with a substantial increase in
favorable SOP votes at the subsequent annual meeting Ferri and Maber (2013) also find that
many firms experiencing low voting dissent at the first SOP vote had removed these practices
before the vote9 highlighting the importance of accounting for ex ante effects when assessing the
impact of a new regulation
Ertimur Ferri and Oesch (2013) examine the effect of SOP on compensation practices in the US
in 2011 (first proxy season under mandatory SOP) Their key findings are generally similar to
the evidence from the UK (i) failed SOP votes are rare (about 2 of the sample) though cases
of substantial dissent are more frequent (ii) more than 55 of the firms experiencing significant
voting dissent respond by making material changes to their compensation plan during the
subsequent year (eg introduction of performance-based vesting conditions in equity grants use
of tougher performance targets removal of perks and tax gross-ups removal of controversial
provisions from severance contracts etc) usually in consultation with major institutional
9 In their sample 700 of low dissent firms shortened their notice periods during the year before the SOP vote
investors (iii) firms making changes to their compensation plans receive greater voting support
at the following SOP vote
Ertimur Ferri and Oesch (2013) also highlight the strong influence of the recommendations
released by proxy advisors (particularly Institutional Shareholder Services ISS) with a negative
recommendation being associated with 25 more votes against the compensation plan Further
evidence of the significant influence of ISS is the striking discontinuity in the relation between
the extent of SOP voting dissent and firmsrsquo responsiveness to the vote the percentage of firms
making compensation changes in response to SOP votes jumps from 32 to 72 around the
30 voting dissent threshold Why After the 2011 proxy season ISS had indicated that firms
failing to ldquoadequatelyrdquo respond to SOP voting dissent above 30 would receive a negative
recommendation in 2012 on the SOP proposal and on the election of compensation committee
members
Finally similar to the UK experience there is also anecdotal evidence of analogous
compensation changes being made by US firms ahead of the SOP vote to avoid a negative
proxy advisor recommendation and an adverse shareholder vote (Thomas Palmiter and Cotter
2013 Larcker McCall and Ormazabal 2013)
III The Effect of Say on Pay on Firm Value
In this section we examine the empirical studies on the overall effect of SOP on firm value
A Event studies around the adoption of Say on Pay regulations
As mentioned in the Introduction on April 20 2007 the House of Representatives passed a SOP
Bill by a 2-1 margin On the same day then-Senator Barack Obama introduced a companion
bill (S1181) in the Senate (which was then put on hold by the Senate Banking Committee)
While the SOP Billrsquos approval was expected (Democrats were in control of the House and
supported the SOP Bill) the 2-1 margin was unexpected suggesting some support for SOP
among Republicans
Cai and Walkling (2011) examine the market reaction to the Housersquos approval of the SOP Bill
for a sample of firms in the SampP 1500 index and document a positive reaction for firms more
likely to benefit from greater shareholder voice over executive pay namely firms with high
abnormal CEO cash pay (defined as the difference between actual CEO cash pay and the level
predicted based on known economic determinants such a size performance industry) firms
with low pay-for-performance sensitivity firms with a history of shareholders willing to vote
against compensation-related management proposals and firms with a history of responsiveness
to shareholder pressure on compensation issues In other words they find a positive reaction in
firms where the compensation problems are more severe and a say on pay vote is more likely to
trigger a response (eg more shareholders willing to vote against management and greater firmsrsquo
propensity to respond to an adverse vote) Note however that they do not find a significant
impact for firms with high abnormal equity and total CEO pay
Larcker Ormazabal and Taylor (2011) also examine the market reaction to the Housersquos approval
of the SOP Bill (in addition to other regulatory events related to executive pay and other
governance provisions) Similar to Cai and Wakling (2011) they fail to find a significant impact
for firms with high abnormal total CEO pay (they do not examine the price reaction for firms
with abnormal cash CEO pay or firms with lower pay-performance sensitivity)
A concern with both studies is that the Housersquos approval of a bill does not guarantee passage in
the Senate and approval by the White House In fact the press at the time reported that the
prospects of the bill in the Senate were uncertain (New York Times 2007) and the Bush White
House openly opposed the Bill (Associated Press 2007) Hence it is not clear the extent to
which this event increased the likelihood of a say on pay legislation Arguably as noted in
Section I the only lsquoeventrsquo that substantially increased the likelihood of say on pay legislation
was the financial crisis and the related perception of compensation abuses (eg outrage over AIG
bonuses)
Ferri and Maber (2013) argue that the announcement of the submission of SOP regulation to the
Parliament in the United Kingdom in June 2002 offers a more powerful setting for an event
study since it was largely unexpected and increased substantially the probability of SOP
adoption (Parliamentrsquos approval was virtually guaranteed) Using this event they document
positive abnormal returns for firms with excess CEO pay combined with poor performance and
for firms with controversial pay practices that weaken the penalties for poor performance (those
practices were often removed in response to SOP votes as discussed earlier in Section IIB)
They interpret their findings as consistent with shareholders perceiving SOP as a value
enhancing monitoring mechanism for firms with low pay-to-poor-performance sensitivity
Finally a recent study by Iliev and Vitanova (2013) examines the market reaction around the
announcement of the SEC final SOP rule that on January 25 2011 exempted small firms (ie
firms with a public float below $75 million) from adopting SOP for two years (press reports at
the time noted that the exemption could become permanent) An appealing feature of this setting
is that the SEC decision was an unexpected reversal from the rule proposed in October 2010 rule
under which no publicly traded firm would be exempted
Iliev and Vitanova (2013) compare the market reaction to this announcement for firms just above
the SEC-imposed threshold and thus subject to the SOP rule and firms just below the threshold
(but otherwise quite similar) and thus exempted from SOP While they find not significant
abnormal returns for either group the abnormal returns for exempted firms are significantly
more negative a finding that the authors interpret as evidence of a positive value effect of
mandatory SOP While the setting is an interesting one some concerns remain as exempted
firms are fairly small and the magnitude of the executive pay problem (if any) is unlikely to be
large enough to explain the differential market reaction Also the same study finds no effect on
CEO pay level and composition (see Section IIA) Hence the reason for the differential stock
price reaction remains unclear It would be helpful to know if the event study result is driven by
(or stronger for) exempted firms with excessive pay or problematic pay practices
B Event studies around Shareholder Proposals to Adopt Say on Pay
Starting in 2006 shareholder activists led by union pension funds submitted shareholder
proposal to adopt SOP at hundreds of US firms in an attempt to induce voluntary or mandatory
widespread adoption of SOP
Cai and Walkling (2011) examine the market reaction to proxy filings and annual meetings of
113 firms targeted by nonbinding shareholder proposals to adopt SOP between 2006 and 2008
On average they find insignificant returns around both the submission of the proposal (proxy
filing date) and the vote on the proposal (annual meeting date) But they also find that when the
proposals are filed by union pension funds the abnormal returns (still insignificant) are more
negative than when filed by other activists and that when the proposal is defeated the abnormal
returns (while insignificant) are more positive than when the proposal is passed (it is not clear
whether this result is driven by union-sponsored proposals) They interpret these findings as
evidence that the market views SOP proposals filed by union pension funds are driven by special
interests rather than value maximization10
However caution is required in interpreting this
evidence Virtually all activists filing SOP proposals were coordinated by a group of investors
(mostly but not only union pension funds) who made available to other activists a template for
SOP proposals and a list of target firms (Ferri and Weber 2009) Hence the distinction between
union and non-union proponents in the context of SOP proposals may be only apparent More
importantly the list of target firms was publicly available months before the proxy filing dates
so it is not clear whether the proxy statements contained any new information Similarly the
voting outcome of the SOP proposals may largely be anticipated (based on the composition of
institutional owners proxy advisorsrsquo recommendations etc) Finally the analyses do not control
for other (potentially new) information contained in proxy statements (eg executive
compensation report other items up for a vote at the annual meeting) or other events occurring at
the annual meeting (eg shareholder votes on other items)
Cuntildeat Gine and Guadalupe (2013) also examine the market reaction to the voting outcome of
nonbinding shareholder proposals to adopt SOP but to alleviate these concerns they employ a
lsquofuzzyrsquo regression discontinuity design (RDD) essentially comparing the stock price reaction to
SOP proposals that pass by a small margin to the reaction to SOP proposals that fail by a small
10
Consistent with this interpretation the authors also find that SOP proposals generally targeted larger firms rather
than firms with excessive pay or lower pay-performance sensitivity However as noted by Ferri and Sandino (2009)
activists typically submit shareholder proposals aimed at promoting policy reforms at a broad sample of large firms
rather than the firms that would most benefit from the proposals because they believe that showing widespread
support for the proposal across all types of firms enhances its credibility with policy makers
margin (Cuntildeat Gine and Guadalupe 2012 Ertimur Ferri and Oesch forthcoming) The underlying idea
is that firms around the threshold are likely to have similar characteristics (indeed they do as the
authors show) but differ in the likelihood of implementation which is typically much higher for
proposals passing the threshold (Ertimur Ferri and Stubben 2010 Thomas and Cotter 2007)
Indeed the authors show that the likelihood of subsequently adopting SOP is 48 higher when
the SOP proposal passes by a small margin relatively to when it fails by a small margin Besides
because in these close-call situations the voting outcome is uncertain its resolution (pass or fail)
is likely to convey new information about the likelihood of SOP adoption making this setting
suitable to an event study11
Using this RDD approach Cuntildeat Gine and Guadalupe (2013) find that on the day of the vote a
SOP proposal that passes by a small margin yields an abnormal return of 24 relative to one
that fails (after controlling for other proposals voted upon at the same meeting) and estimate the
lsquofullrsquo value of SOP at 46 (after taking into account the increase in the probability of SOP
implementation)12
Aside from the issues discussed in Section IA (ie generalizability to other
firms) this estimate seems too large to reflect the present value of future reductions in excess
CEO pay (as noted by the authors their estimate of the value of SOP is equivalent to the
estimated value of removing two anti-takeover provision based on Cuntildeat Gine and Guadalupe
2012) particularly because (i) the authors find no evidence of subsequent changes in levels and
composition of CEO pay for firms implementing SOP (see Section IIA) an (ii) the sample firms
11
Consistent with this observation the authors find no market reaction to the voting outcome of SOP shareholder
proposals away from the threshold
12 Since the outcome of the vote is not binding the 24 market reaction only reflects the expected increase in the
probability of SOP adoption and thus understates the value of the SOP provision
do not appear to be characterized by excess CEO pay However the authors also find significant
improvements in subsequent operating performance and efficiency as a result of passing SOP
proposals and thus conjecture that the SOP vote is viewed as a tool to express a vote of
confidence in management performance more than a way to change compensation practices and
that the large positive market reaction reflects expected performance improvements under this
tighter monitoring regime While this is an interesting idea it remains unclear what additional
ldquoteethrdquo a SOP vote may provide over existing (non-compensation related) monitoring
mechanisms shareholders can express (and do often express) their dissatisfaction with
management performance when voting on director elections (Cai Garner and Walkling 2009)
C Event studies around compensation changes induced by SOP
Two studies examine announcements of changes to compensation plans related to SOP Larcker
McCall and Ormazabal (2013) examine the market reaction to (non-contaminated) compensation
changes disclosed in 8-K filings by Russell 3000 firms in the US during the year before the first
SOP vote They find that changes that appear to be made to avoid a negative proxy advisorrsquos
recommendation and thus a negative SOP vote are associated with a negative abnormal return
of -044 whereas other compensation changes are not associated with a significant stock price
reaction Ertimur Ferri and Oesch (2013) perform a similar analysis but focusing on
compensation changes made after the first SOP vote (and explicitly in response to the vote) by
firms receiving a negative recommendation and a high voting dissent in 2011 They fail to find
significant abnormal returns even for the subset of compensation changes that resulted in a
positive recommendation and low dissent in 2012 (and thus were presumably perceived to be
adequate and material by proxy advisors and voting shareholders)
D Other approaches effect of SOP on Tobinrsquos Q
In their cross-country study Correa and Lel (2013) also examine the effect of SOP laws on
Tobinrsquos Q and report a 36 increase in firm value following the adoption of the SOP laws
(more precisely a 36 higher firm value for post-SOP observations relative to a control sample
that pools together pre-SOP and non-SOP observations) They acknowledge that this increase in
firm value is too large to be justified by the relative decrease in CEO pay that they document and
suggest (but do not test) that it may reflect better alignment of pay and performance13
A key
concern is that higher valuation in countries with SOP laws may reflect other governance
changes introduced at the same time While the study controls for other compensation-related
laws there may be other non-compensation regulations introduced with SOP and with a
potentially larger impact on firm value
IV What have we learned
Adoption of SOP in the US has been long advocated by those who contend that executive pay is
a manifestation of rather than a solution to the agency problem and by those who favor great
shareholder involvement in corporate decisions A growing body of research is starting to
investigate the effect of SOP on executive pay and firm value It is premature to draw definitive
conclusions also because these studies differ in methodologies and settings but three points
seem to emerge First there is robust evidence that boards respond to SOP votes when
13
They also find that the firm value increase is higher when the relative decrease in CEO pay results in a lower pay
differential between CEO and other top managers (CEO pay slice) implying that a source of value creation may be
the reduced pay inequality among the top management team But this seems unlikely to explain the change in
Tobinrsquos Q since the increase occurs for both firms with advisory SOP laws and firms with binding SOP laws and
there is no change in CEO pay slice in firms subject to binding SOP laws
shareholders use it ldquovoicerdquo is heardrdquo Both in the US and UK studies show that not only firms
failing to win the SOP vote but also firms facing substantial dissent (20-30 of the votes
against) make changes to their compensation plans in consultation with institutional investors
and proxy advisors Second in both the US and UK institutional investors appear to use the
power of SOP votes to pressure firms into strengthening the link (or the perceived link) between
pay and performance (particularly on the downside) but not to pressure firms to reduce target
levels of pay suggesting a reluctance of institutional investors to lsquoregulatersquo executive pay Third
and consistent with the above most studies examining the aggregate effect of SOP on CEO pay
find some increase in pay-performance sensitivity but not effect on the level and growth of CEO
pay While some observers may interpret the lack of an effect on CEO pay levels as evidence of
the failure of SOP one should note that SOP per se is a neutral tool Its impact depends on what
investors choose to lsquosay on payrsquo Fourth based on my interpretation of the existing research to
date there is little evidence of an effect of SOP on firm value The studies documenting a
positive price impact appear subject to alternative interpretations or not plausible in their
findings (eg they fail to find an effect on CEO pay or the effect is too small to justify the
documented price impact hence it remains unclear what the source of value creation is) Also
there seems to be no stock price reaction to the SOP-induced compensation changes Finally
given the nature of these SOP-induced changes while some of them may be beneficial it seems
hard to believe that they will have a statistically detectable impact on firm value except perhaps
in a handful of firms where the compensation plan is subject to a complete overhaul
Perhaps one explanation for the weak impact of SOP is that executive pay problems were
overstated or that by the time SOP was introduced (more than a decade after the Enron-type
scandals that led to calls for greater shareholder voice) they had been already addressed via
other mechanisms (hedge fund activism monitoring by institutional investors vote-no
campaigns against compensation committee members SEC-mandated pay disclosures)
Overall though based on the evidence to date it does not appear that SOP had a major impact14
Was the adoption of SOP an lsquooptimalrsquo choice from a social welfare point of view This is a
difficult question to answer given the challenges of identifying and measuring all the potential
costs and benefits associated with SOP (likely to change over time and differ across countries)
Perhaps the most positive view of SOP is that its introduction prevented the adoption of other
more radical and intrusive regulatory measures (eg CEO pay caps) In that sense it may be
viewed as an lsquooptimalrsquo answer (ie the lesser of two evils) to the political pressure to reform
executive pay during the financial crisis
14
This interpretation of the evidence is also consistent with the prediction of analytical studies Ozbas and
Matsusaka (2013) model the benefits and costs of shareholder empowerment distinguishing between the right to
approve and the right to propose They show that permitting shareholders to propose directors or policies can cause
value-maximizing managers to take value-reducing actions to accommodate activist investors with non-value-
maximizing goals As for the right to approve it is weakly beneficial that is approval rights (such as a SOP vote)
are generally beneficial for shareholders in that they limit the managerrsquos ability to pursue private benefits at
shareholder expense but they have minimal impact on managerial actions and firm value (because the manager in
effect can threaten shareholders with an undesirable status quo if they do not approve the managerrsquos proposed
action)
References
Associated Press 2007 Administration opposes say on pay bill April 19
Bainbridge S 2008 Remarks on say on pay An unjustified incursion on director authority
UCLA School of Law Research Paper No 08-06
Bebchuk L 2005 The Case for Increasing Shareholder Power Harvard Law Review 118833-
917
Bebchuk L 2007 Written testimony submitted before the Committee on Financial Services
United States House of Representatives Hearing on Empowering Shareholders on Executive
Compensation March 8
Bebchuk LA A Cohen and A Ferrell 2009 What Matters in Corporate Governance Review
of Financial Studies 22783-827
Cai J J Garner and R Walkling 2009 Electing Directors Journal of Finance 642387-2419
Cai J and R Walkling 2011 Shareholdersrsquo Say on Pay Does it Create Value Journal of
Financial and Quantitative Analysis 46299ndash339
Coates IV JC (2009) Testimony before the Subcommittee on Securities Insurance and
Investment of the Committee on Banking Housing and Urban Affairs United States Senate
July 29 2009 httppapersssrncomsol3paperscfmabstract_id=1475355
Correa R and U Lel 2013 Say On Pay Laws Executive Compensation CEO Pay Slice and
Firm Value Around the World Federal Reserve Board Working Paper
Cuntildeat V M Gine and M Guadalupe 2012 The Vote is Cast The Effect of Corporate
Governance on Shareholder Value Journal of Finance 671943-1977
Cuntildeat V M Gine and M Guadalupe 2013 Say Pays Shareholder Voice and Firm
Performance ECGI - Finance Working Paper No 373
Del Guercio D L Seery and T Woidtke 2008 Do Boards Pay Attention When Institutional
Investors lsquoJust Vote Norsquo Journal of Financial Economics 9084-103
Ertimur Y F Ferri and S Stubben 2010 Board of Directorslsquo Responsiveness to Shareholders
Evidence from Shareholder Proposals Journal of Corporate Finance 1653-72
Ertimur Y F Ferri and V Muslu 2011 Shareholder Activism and CEO Pay Review of
Financial Studies 24535ndash592
Ertimur Y F Ferri and D Oesch 2013 Shareholder Votes and Proxy Advisors Evidence from
Say on Pay Journal of Accounting Research 51951-996
Ertimur Y F Ferri and D Oesch forthcoming Does the Director Election System Matter
Evidence from Majority Voting Review of Accounting Studies forthcoming
Ferri F 2012 Low-Cost Shareholder Activism A Review of the Evidence Research
Handbook on the Economics of Corporate Law ed CA Hill and BH McDonnell
Edward Elgar Publishing
Ferri F and D Maber 2013 Say on Pay Votes and CEO Compensation Evidence from the UK
Review of Finance 17527-563
Ferri F and D Oesch 2013 Management Influence on Investors Evidence from Shareholder
Votes on the Frequency of Say on Pay Columbia Business School Research Paper No 13-17
Ferri F and T Sandino 2009 The impact of shareholder activism on financial reporting and
compensation The case of employee stock options expensing The Accounting Review 84433ndash
466
Ferri F and J Weber 2009 AFSCME vs Mozilohellipand Say on Pay for All (A) Harvard
Business School Case 109-009
Gompers P J Ishii and A Metrick 2003 Corporate Governance and Equity Prices Quarterly
Journal of Economics 118107-155
Gordon J 2009 Say on Paylsquo Cautionary Notes on the UK Experience and the Case for
Shareholders Opt-In Harvard Journal on Legislation 46323-368
Iliev P and S Vitanova 2013 The effect of the Say on Pay in the US Pennsylvania State
University Working Paper
Kaplan S 2007 Testimony of Steven N Kaplan on Empowering Shareholders on Executive
Compensation and HR 1257 the Shareholder Vote on Executive Compensation Act before the
Committee on Financial Services United States House of Representatives March 8
Larcker D A McCall and G Ormazabal 2013 The Economic Consequences of Proxy
Advisor Say-on-Pay Voting Policies Working Paper Stanford University
Larcker D G Ormazabal and D Taylor 2011 The market reaction to corporate governance
regulation Journal of Financial Economics 101431ndash448
Levit D and N Malenko 2011 Non-binding voting for shareholder proposals Journal of
Finance 661579ndash1614
Mishel L and N Sabadish 2012 How executive compensation and financial-sector pay have
fueled income inequality Issue Brief 331 Economic Policy Institute
Murphy K 2012 The Politics of Pay A Legislative History of Executive Compensation in
Jennifer Hill and Randall Thomas eds Research Handbook on Executive Pay Edward Elgar
Publishers
New York Times 2007 House votes to give investors say on executive pay April 21
Norris F 2004 Corporate democracy and the power to embarrass New York Times March 4
Ozbas O and J Matsusaka2013 Managerial Accommodation Proxy Access and the Cost of
Shareholder Empowerment University of Southern California CLEO Research Paper Series No
C12-1
Thomas R and J Cotter 2007 Shareholder proposals in the new millennium Shareholder
support board response and market reaction Journal of Corporate Finance 13 368ndash391
Thomas R A Palmiter and J Cotter 2012 Dodd-Franks Say on Pay Will it Lead to a Greater
Role for Shareholders in Corporate Governance Cornell Law Review 971213-1266
Thomas R and C Van der Elst 2013 The International Scope of Say on Pay ECGI Law
Working Paper No227
not be surprising More generally the findings cannot be necessarily generalized to firms away
from the 50 threshold or to firms not targeted by SOP shareholder proposals
B Effect of SOP votes on compensation practices
Analyses of the effect of SOP on level and composition of CEO pay may not capture a number
of other changes to compensation contracts that may be induced by SOP votes but are not
reflected in measure of CEO pay For example the introduction of performance-based vesting
provisions is typically not reflected in the estimates of fair value of equity grants used by
researchers in measuring CEO pay Many other changes (eg terms of severance packages) will
only be reflected in measures of CEO pay contingent upon the occurrence of certain events
Hence it is important to complement the evidence in section IIA with an analysis of firmsrsquo
disclosures of changes to observable provisions of compensation contracts Another benefit of
this approach is that it captures the changes that boards explicitly present as a response to SOP
votes and thus it acts as a reality check on regression-based inferences For example a
regression-based finding of say a decrease in CEO pay level would be more credible if
supported by evidence that firms disclose a decision to reduce target levels of CEO pay in
response to an adverse SOP vote (an action that presumably they have an incentive to disclose to
get rewarded by shareholders in subsequent votes)
Two studies collect data on specific changes made to compensation contracts explicitly in
response to SOP votes in the UK and the US using firmsrsquo disclosures in their proxy statements
Ferri and Maber (2013) examine the compensation reports of a sample of UK firms before and
after the first SOP vote and find that firms experiencing higher voting dissent were significantly
more likely to remove compensation provisions criticized by investors as lsquorewards for failurersquo
relative to a matched sample of firms experiencing lower dissent For example they report that
among firms with long notice periods (implying larger severance payments ) the percentage of
high dissent firms that shortened them after the vote (800) was significantly higher than before
the vote (200) and also significantly higher than among low dissent firms after the vote
(333) They report similar findings for another controversial practice namely the presence of
retesting provisions in the performance-based vesting conditions of equity grants Most firms
indicate that these changes were the result of consultations with their major institutional
investors Firmsrsquo responsiveness to SOP votes was lsquorewardedrsquo with a substantial increase in
favorable SOP votes at the subsequent annual meeting Ferri and Maber (2013) also find that
many firms experiencing low voting dissent at the first SOP vote had removed these practices
before the vote9 highlighting the importance of accounting for ex ante effects when assessing the
impact of a new regulation
Ertimur Ferri and Oesch (2013) examine the effect of SOP on compensation practices in the US
in 2011 (first proxy season under mandatory SOP) Their key findings are generally similar to
the evidence from the UK (i) failed SOP votes are rare (about 2 of the sample) though cases
of substantial dissent are more frequent (ii) more than 55 of the firms experiencing significant
voting dissent respond by making material changes to their compensation plan during the
subsequent year (eg introduction of performance-based vesting conditions in equity grants use
of tougher performance targets removal of perks and tax gross-ups removal of controversial
provisions from severance contracts etc) usually in consultation with major institutional
9 In their sample 700 of low dissent firms shortened their notice periods during the year before the SOP vote
investors (iii) firms making changes to their compensation plans receive greater voting support
at the following SOP vote
Ertimur Ferri and Oesch (2013) also highlight the strong influence of the recommendations
released by proxy advisors (particularly Institutional Shareholder Services ISS) with a negative
recommendation being associated with 25 more votes against the compensation plan Further
evidence of the significant influence of ISS is the striking discontinuity in the relation between
the extent of SOP voting dissent and firmsrsquo responsiveness to the vote the percentage of firms
making compensation changes in response to SOP votes jumps from 32 to 72 around the
30 voting dissent threshold Why After the 2011 proxy season ISS had indicated that firms
failing to ldquoadequatelyrdquo respond to SOP voting dissent above 30 would receive a negative
recommendation in 2012 on the SOP proposal and on the election of compensation committee
members
Finally similar to the UK experience there is also anecdotal evidence of analogous
compensation changes being made by US firms ahead of the SOP vote to avoid a negative
proxy advisor recommendation and an adverse shareholder vote (Thomas Palmiter and Cotter
2013 Larcker McCall and Ormazabal 2013)
III The Effect of Say on Pay on Firm Value
In this section we examine the empirical studies on the overall effect of SOP on firm value
A Event studies around the adoption of Say on Pay regulations
As mentioned in the Introduction on April 20 2007 the House of Representatives passed a SOP
Bill by a 2-1 margin On the same day then-Senator Barack Obama introduced a companion
bill (S1181) in the Senate (which was then put on hold by the Senate Banking Committee)
While the SOP Billrsquos approval was expected (Democrats were in control of the House and
supported the SOP Bill) the 2-1 margin was unexpected suggesting some support for SOP
among Republicans
Cai and Walkling (2011) examine the market reaction to the Housersquos approval of the SOP Bill
for a sample of firms in the SampP 1500 index and document a positive reaction for firms more
likely to benefit from greater shareholder voice over executive pay namely firms with high
abnormal CEO cash pay (defined as the difference between actual CEO cash pay and the level
predicted based on known economic determinants such a size performance industry) firms
with low pay-for-performance sensitivity firms with a history of shareholders willing to vote
against compensation-related management proposals and firms with a history of responsiveness
to shareholder pressure on compensation issues In other words they find a positive reaction in
firms where the compensation problems are more severe and a say on pay vote is more likely to
trigger a response (eg more shareholders willing to vote against management and greater firmsrsquo
propensity to respond to an adverse vote) Note however that they do not find a significant
impact for firms with high abnormal equity and total CEO pay
Larcker Ormazabal and Taylor (2011) also examine the market reaction to the Housersquos approval
of the SOP Bill (in addition to other regulatory events related to executive pay and other
governance provisions) Similar to Cai and Wakling (2011) they fail to find a significant impact
for firms with high abnormal total CEO pay (they do not examine the price reaction for firms
with abnormal cash CEO pay or firms with lower pay-performance sensitivity)
A concern with both studies is that the Housersquos approval of a bill does not guarantee passage in
the Senate and approval by the White House In fact the press at the time reported that the
prospects of the bill in the Senate were uncertain (New York Times 2007) and the Bush White
House openly opposed the Bill (Associated Press 2007) Hence it is not clear the extent to
which this event increased the likelihood of a say on pay legislation Arguably as noted in
Section I the only lsquoeventrsquo that substantially increased the likelihood of say on pay legislation
was the financial crisis and the related perception of compensation abuses (eg outrage over AIG
bonuses)
Ferri and Maber (2013) argue that the announcement of the submission of SOP regulation to the
Parliament in the United Kingdom in June 2002 offers a more powerful setting for an event
study since it was largely unexpected and increased substantially the probability of SOP
adoption (Parliamentrsquos approval was virtually guaranteed) Using this event they document
positive abnormal returns for firms with excess CEO pay combined with poor performance and
for firms with controversial pay practices that weaken the penalties for poor performance (those
practices were often removed in response to SOP votes as discussed earlier in Section IIB)
They interpret their findings as consistent with shareholders perceiving SOP as a value
enhancing monitoring mechanism for firms with low pay-to-poor-performance sensitivity
Finally a recent study by Iliev and Vitanova (2013) examines the market reaction around the
announcement of the SEC final SOP rule that on January 25 2011 exempted small firms (ie
firms with a public float below $75 million) from adopting SOP for two years (press reports at
the time noted that the exemption could become permanent) An appealing feature of this setting
is that the SEC decision was an unexpected reversal from the rule proposed in October 2010 rule
under which no publicly traded firm would be exempted
Iliev and Vitanova (2013) compare the market reaction to this announcement for firms just above
the SEC-imposed threshold and thus subject to the SOP rule and firms just below the threshold
(but otherwise quite similar) and thus exempted from SOP While they find not significant
abnormal returns for either group the abnormal returns for exempted firms are significantly
more negative a finding that the authors interpret as evidence of a positive value effect of
mandatory SOP While the setting is an interesting one some concerns remain as exempted
firms are fairly small and the magnitude of the executive pay problem (if any) is unlikely to be
large enough to explain the differential market reaction Also the same study finds no effect on
CEO pay level and composition (see Section IIA) Hence the reason for the differential stock
price reaction remains unclear It would be helpful to know if the event study result is driven by
(or stronger for) exempted firms with excessive pay or problematic pay practices
B Event studies around Shareholder Proposals to Adopt Say on Pay
Starting in 2006 shareholder activists led by union pension funds submitted shareholder
proposal to adopt SOP at hundreds of US firms in an attempt to induce voluntary or mandatory
widespread adoption of SOP
Cai and Walkling (2011) examine the market reaction to proxy filings and annual meetings of
113 firms targeted by nonbinding shareholder proposals to adopt SOP between 2006 and 2008
On average they find insignificant returns around both the submission of the proposal (proxy
filing date) and the vote on the proposal (annual meeting date) But they also find that when the
proposals are filed by union pension funds the abnormal returns (still insignificant) are more
negative than when filed by other activists and that when the proposal is defeated the abnormal
returns (while insignificant) are more positive than when the proposal is passed (it is not clear
whether this result is driven by union-sponsored proposals) They interpret these findings as
evidence that the market views SOP proposals filed by union pension funds are driven by special
interests rather than value maximization10
However caution is required in interpreting this
evidence Virtually all activists filing SOP proposals were coordinated by a group of investors
(mostly but not only union pension funds) who made available to other activists a template for
SOP proposals and a list of target firms (Ferri and Weber 2009) Hence the distinction between
union and non-union proponents in the context of SOP proposals may be only apparent More
importantly the list of target firms was publicly available months before the proxy filing dates
so it is not clear whether the proxy statements contained any new information Similarly the
voting outcome of the SOP proposals may largely be anticipated (based on the composition of
institutional owners proxy advisorsrsquo recommendations etc) Finally the analyses do not control
for other (potentially new) information contained in proxy statements (eg executive
compensation report other items up for a vote at the annual meeting) or other events occurring at
the annual meeting (eg shareholder votes on other items)
Cuntildeat Gine and Guadalupe (2013) also examine the market reaction to the voting outcome of
nonbinding shareholder proposals to adopt SOP but to alleviate these concerns they employ a
lsquofuzzyrsquo regression discontinuity design (RDD) essentially comparing the stock price reaction to
SOP proposals that pass by a small margin to the reaction to SOP proposals that fail by a small
10
Consistent with this interpretation the authors also find that SOP proposals generally targeted larger firms rather
than firms with excessive pay or lower pay-performance sensitivity However as noted by Ferri and Sandino (2009)
activists typically submit shareholder proposals aimed at promoting policy reforms at a broad sample of large firms
rather than the firms that would most benefit from the proposals because they believe that showing widespread
support for the proposal across all types of firms enhances its credibility with policy makers
margin (Cuntildeat Gine and Guadalupe 2012 Ertimur Ferri and Oesch forthcoming) The underlying idea
is that firms around the threshold are likely to have similar characteristics (indeed they do as the
authors show) but differ in the likelihood of implementation which is typically much higher for
proposals passing the threshold (Ertimur Ferri and Stubben 2010 Thomas and Cotter 2007)
Indeed the authors show that the likelihood of subsequently adopting SOP is 48 higher when
the SOP proposal passes by a small margin relatively to when it fails by a small margin Besides
because in these close-call situations the voting outcome is uncertain its resolution (pass or fail)
is likely to convey new information about the likelihood of SOP adoption making this setting
suitable to an event study11
Using this RDD approach Cuntildeat Gine and Guadalupe (2013) find that on the day of the vote a
SOP proposal that passes by a small margin yields an abnormal return of 24 relative to one
that fails (after controlling for other proposals voted upon at the same meeting) and estimate the
lsquofullrsquo value of SOP at 46 (after taking into account the increase in the probability of SOP
implementation)12
Aside from the issues discussed in Section IA (ie generalizability to other
firms) this estimate seems too large to reflect the present value of future reductions in excess
CEO pay (as noted by the authors their estimate of the value of SOP is equivalent to the
estimated value of removing two anti-takeover provision based on Cuntildeat Gine and Guadalupe
2012) particularly because (i) the authors find no evidence of subsequent changes in levels and
composition of CEO pay for firms implementing SOP (see Section IIA) an (ii) the sample firms
11
Consistent with this observation the authors find no market reaction to the voting outcome of SOP shareholder
proposals away from the threshold
12 Since the outcome of the vote is not binding the 24 market reaction only reflects the expected increase in the
probability of SOP adoption and thus understates the value of the SOP provision
do not appear to be characterized by excess CEO pay However the authors also find significant
improvements in subsequent operating performance and efficiency as a result of passing SOP
proposals and thus conjecture that the SOP vote is viewed as a tool to express a vote of
confidence in management performance more than a way to change compensation practices and
that the large positive market reaction reflects expected performance improvements under this
tighter monitoring regime While this is an interesting idea it remains unclear what additional
ldquoteethrdquo a SOP vote may provide over existing (non-compensation related) monitoring
mechanisms shareholders can express (and do often express) their dissatisfaction with
management performance when voting on director elections (Cai Garner and Walkling 2009)
C Event studies around compensation changes induced by SOP
Two studies examine announcements of changes to compensation plans related to SOP Larcker
McCall and Ormazabal (2013) examine the market reaction to (non-contaminated) compensation
changes disclosed in 8-K filings by Russell 3000 firms in the US during the year before the first
SOP vote They find that changes that appear to be made to avoid a negative proxy advisorrsquos
recommendation and thus a negative SOP vote are associated with a negative abnormal return
of -044 whereas other compensation changes are not associated with a significant stock price
reaction Ertimur Ferri and Oesch (2013) perform a similar analysis but focusing on
compensation changes made after the first SOP vote (and explicitly in response to the vote) by
firms receiving a negative recommendation and a high voting dissent in 2011 They fail to find
significant abnormal returns even for the subset of compensation changes that resulted in a
positive recommendation and low dissent in 2012 (and thus were presumably perceived to be
adequate and material by proxy advisors and voting shareholders)
D Other approaches effect of SOP on Tobinrsquos Q
In their cross-country study Correa and Lel (2013) also examine the effect of SOP laws on
Tobinrsquos Q and report a 36 increase in firm value following the adoption of the SOP laws
(more precisely a 36 higher firm value for post-SOP observations relative to a control sample
that pools together pre-SOP and non-SOP observations) They acknowledge that this increase in
firm value is too large to be justified by the relative decrease in CEO pay that they document and
suggest (but do not test) that it may reflect better alignment of pay and performance13
A key
concern is that higher valuation in countries with SOP laws may reflect other governance
changes introduced at the same time While the study controls for other compensation-related
laws there may be other non-compensation regulations introduced with SOP and with a
potentially larger impact on firm value
IV What have we learned
Adoption of SOP in the US has been long advocated by those who contend that executive pay is
a manifestation of rather than a solution to the agency problem and by those who favor great
shareholder involvement in corporate decisions A growing body of research is starting to
investigate the effect of SOP on executive pay and firm value It is premature to draw definitive
conclusions also because these studies differ in methodologies and settings but three points
seem to emerge First there is robust evidence that boards respond to SOP votes when
13
They also find that the firm value increase is higher when the relative decrease in CEO pay results in a lower pay
differential between CEO and other top managers (CEO pay slice) implying that a source of value creation may be
the reduced pay inequality among the top management team But this seems unlikely to explain the change in
Tobinrsquos Q since the increase occurs for both firms with advisory SOP laws and firms with binding SOP laws and
there is no change in CEO pay slice in firms subject to binding SOP laws
shareholders use it ldquovoicerdquo is heardrdquo Both in the US and UK studies show that not only firms
failing to win the SOP vote but also firms facing substantial dissent (20-30 of the votes
against) make changes to their compensation plans in consultation with institutional investors
and proxy advisors Second in both the US and UK institutional investors appear to use the
power of SOP votes to pressure firms into strengthening the link (or the perceived link) between
pay and performance (particularly on the downside) but not to pressure firms to reduce target
levels of pay suggesting a reluctance of institutional investors to lsquoregulatersquo executive pay Third
and consistent with the above most studies examining the aggregate effect of SOP on CEO pay
find some increase in pay-performance sensitivity but not effect on the level and growth of CEO
pay While some observers may interpret the lack of an effect on CEO pay levels as evidence of
the failure of SOP one should note that SOP per se is a neutral tool Its impact depends on what
investors choose to lsquosay on payrsquo Fourth based on my interpretation of the existing research to
date there is little evidence of an effect of SOP on firm value The studies documenting a
positive price impact appear subject to alternative interpretations or not plausible in their
findings (eg they fail to find an effect on CEO pay or the effect is too small to justify the
documented price impact hence it remains unclear what the source of value creation is) Also
there seems to be no stock price reaction to the SOP-induced compensation changes Finally
given the nature of these SOP-induced changes while some of them may be beneficial it seems
hard to believe that they will have a statistically detectable impact on firm value except perhaps
in a handful of firms where the compensation plan is subject to a complete overhaul
Perhaps one explanation for the weak impact of SOP is that executive pay problems were
overstated or that by the time SOP was introduced (more than a decade after the Enron-type
scandals that led to calls for greater shareholder voice) they had been already addressed via
other mechanisms (hedge fund activism monitoring by institutional investors vote-no
campaigns against compensation committee members SEC-mandated pay disclosures)
Overall though based on the evidence to date it does not appear that SOP had a major impact14
Was the adoption of SOP an lsquooptimalrsquo choice from a social welfare point of view This is a
difficult question to answer given the challenges of identifying and measuring all the potential
costs and benefits associated with SOP (likely to change over time and differ across countries)
Perhaps the most positive view of SOP is that its introduction prevented the adoption of other
more radical and intrusive regulatory measures (eg CEO pay caps) In that sense it may be
viewed as an lsquooptimalrsquo answer (ie the lesser of two evils) to the political pressure to reform
executive pay during the financial crisis
14
This interpretation of the evidence is also consistent with the prediction of analytical studies Ozbas and
Matsusaka (2013) model the benefits and costs of shareholder empowerment distinguishing between the right to
approve and the right to propose They show that permitting shareholders to propose directors or policies can cause
value-maximizing managers to take value-reducing actions to accommodate activist investors with non-value-
maximizing goals As for the right to approve it is weakly beneficial that is approval rights (such as a SOP vote)
are generally beneficial for shareholders in that they limit the managerrsquos ability to pursue private benefits at
shareholder expense but they have minimal impact on managerial actions and firm value (because the manager in
effect can threaten shareholders with an undesirable status quo if they do not approve the managerrsquos proposed
action)
References
Associated Press 2007 Administration opposes say on pay bill April 19
Bainbridge S 2008 Remarks on say on pay An unjustified incursion on director authority
UCLA School of Law Research Paper No 08-06
Bebchuk L 2005 The Case for Increasing Shareholder Power Harvard Law Review 118833-
917
Bebchuk L 2007 Written testimony submitted before the Committee on Financial Services
United States House of Representatives Hearing on Empowering Shareholders on Executive
Compensation March 8
Bebchuk LA A Cohen and A Ferrell 2009 What Matters in Corporate Governance Review
of Financial Studies 22783-827
Cai J J Garner and R Walkling 2009 Electing Directors Journal of Finance 642387-2419
Cai J and R Walkling 2011 Shareholdersrsquo Say on Pay Does it Create Value Journal of
Financial and Quantitative Analysis 46299ndash339
Coates IV JC (2009) Testimony before the Subcommittee on Securities Insurance and
Investment of the Committee on Banking Housing and Urban Affairs United States Senate
July 29 2009 httppapersssrncomsol3paperscfmabstract_id=1475355
Correa R and U Lel 2013 Say On Pay Laws Executive Compensation CEO Pay Slice and
Firm Value Around the World Federal Reserve Board Working Paper
Cuntildeat V M Gine and M Guadalupe 2012 The Vote is Cast The Effect of Corporate
Governance on Shareholder Value Journal of Finance 671943-1977
Cuntildeat V M Gine and M Guadalupe 2013 Say Pays Shareholder Voice and Firm
Performance ECGI - Finance Working Paper No 373
Del Guercio D L Seery and T Woidtke 2008 Do Boards Pay Attention When Institutional
Investors lsquoJust Vote Norsquo Journal of Financial Economics 9084-103
Ertimur Y F Ferri and S Stubben 2010 Board of Directorslsquo Responsiveness to Shareholders
Evidence from Shareholder Proposals Journal of Corporate Finance 1653-72
Ertimur Y F Ferri and V Muslu 2011 Shareholder Activism and CEO Pay Review of
Financial Studies 24535ndash592
Ertimur Y F Ferri and D Oesch 2013 Shareholder Votes and Proxy Advisors Evidence from
Say on Pay Journal of Accounting Research 51951-996
Ertimur Y F Ferri and D Oesch forthcoming Does the Director Election System Matter
Evidence from Majority Voting Review of Accounting Studies forthcoming
Ferri F 2012 Low-Cost Shareholder Activism A Review of the Evidence Research
Handbook on the Economics of Corporate Law ed CA Hill and BH McDonnell
Edward Elgar Publishing
Ferri F and D Maber 2013 Say on Pay Votes and CEO Compensation Evidence from the UK
Review of Finance 17527-563
Ferri F and D Oesch 2013 Management Influence on Investors Evidence from Shareholder
Votes on the Frequency of Say on Pay Columbia Business School Research Paper No 13-17
Ferri F and T Sandino 2009 The impact of shareholder activism on financial reporting and
compensation The case of employee stock options expensing The Accounting Review 84433ndash
466
Ferri F and J Weber 2009 AFSCME vs Mozilohellipand Say on Pay for All (A) Harvard
Business School Case 109-009
Gompers P J Ishii and A Metrick 2003 Corporate Governance and Equity Prices Quarterly
Journal of Economics 118107-155
Gordon J 2009 Say on Paylsquo Cautionary Notes on the UK Experience and the Case for
Shareholders Opt-In Harvard Journal on Legislation 46323-368
Iliev P and S Vitanova 2013 The effect of the Say on Pay in the US Pennsylvania State
University Working Paper
Kaplan S 2007 Testimony of Steven N Kaplan on Empowering Shareholders on Executive
Compensation and HR 1257 the Shareholder Vote on Executive Compensation Act before the
Committee on Financial Services United States House of Representatives March 8
Larcker D A McCall and G Ormazabal 2013 The Economic Consequences of Proxy
Advisor Say-on-Pay Voting Policies Working Paper Stanford University
Larcker D G Ormazabal and D Taylor 2011 The market reaction to corporate governance
regulation Journal of Financial Economics 101431ndash448
Levit D and N Malenko 2011 Non-binding voting for shareholder proposals Journal of
Finance 661579ndash1614
Mishel L and N Sabadish 2012 How executive compensation and financial-sector pay have
fueled income inequality Issue Brief 331 Economic Policy Institute
Murphy K 2012 The Politics of Pay A Legislative History of Executive Compensation in
Jennifer Hill and Randall Thomas eds Research Handbook on Executive Pay Edward Elgar
Publishers
New York Times 2007 House votes to give investors say on executive pay April 21
Norris F 2004 Corporate democracy and the power to embarrass New York Times March 4
Ozbas O and J Matsusaka2013 Managerial Accommodation Proxy Access and the Cost of
Shareholder Empowerment University of Southern California CLEO Research Paper Series No
C12-1
Thomas R and J Cotter 2007 Shareholder proposals in the new millennium Shareholder
support board response and market reaction Journal of Corporate Finance 13 368ndash391
Thomas R A Palmiter and J Cotter 2012 Dodd-Franks Say on Pay Will it Lead to a Greater
Role for Shareholders in Corporate Governance Cornell Law Review 971213-1266
Thomas R and C Van der Elst 2013 The International Scope of Say on Pay ECGI Law
Working Paper No227
relative to a matched sample of firms experiencing lower dissent For example they report that
among firms with long notice periods (implying larger severance payments ) the percentage of
high dissent firms that shortened them after the vote (800) was significantly higher than before
the vote (200) and also significantly higher than among low dissent firms after the vote
(333) They report similar findings for another controversial practice namely the presence of
retesting provisions in the performance-based vesting conditions of equity grants Most firms
indicate that these changes were the result of consultations with their major institutional
investors Firmsrsquo responsiveness to SOP votes was lsquorewardedrsquo with a substantial increase in
favorable SOP votes at the subsequent annual meeting Ferri and Maber (2013) also find that
many firms experiencing low voting dissent at the first SOP vote had removed these practices
before the vote9 highlighting the importance of accounting for ex ante effects when assessing the
impact of a new regulation
Ertimur Ferri and Oesch (2013) examine the effect of SOP on compensation practices in the US
in 2011 (first proxy season under mandatory SOP) Their key findings are generally similar to
the evidence from the UK (i) failed SOP votes are rare (about 2 of the sample) though cases
of substantial dissent are more frequent (ii) more than 55 of the firms experiencing significant
voting dissent respond by making material changes to their compensation plan during the
subsequent year (eg introduction of performance-based vesting conditions in equity grants use
of tougher performance targets removal of perks and tax gross-ups removal of controversial
provisions from severance contracts etc) usually in consultation with major institutional
9 In their sample 700 of low dissent firms shortened their notice periods during the year before the SOP vote
investors (iii) firms making changes to their compensation plans receive greater voting support
at the following SOP vote
Ertimur Ferri and Oesch (2013) also highlight the strong influence of the recommendations
released by proxy advisors (particularly Institutional Shareholder Services ISS) with a negative
recommendation being associated with 25 more votes against the compensation plan Further
evidence of the significant influence of ISS is the striking discontinuity in the relation between
the extent of SOP voting dissent and firmsrsquo responsiveness to the vote the percentage of firms
making compensation changes in response to SOP votes jumps from 32 to 72 around the
30 voting dissent threshold Why After the 2011 proxy season ISS had indicated that firms
failing to ldquoadequatelyrdquo respond to SOP voting dissent above 30 would receive a negative
recommendation in 2012 on the SOP proposal and on the election of compensation committee
members
Finally similar to the UK experience there is also anecdotal evidence of analogous
compensation changes being made by US firms ahead of the SOP vote to avoid a negative
proxy advisor recommendation and an adverse shareholder vote (Thomas Palmiter and Cotter
2013 Larcker McCall and Ormazabal 2013)
III The Effect of Say on Pay on Firm Value
In this section we examine the empirical studies on the overall effect of SOP on firm value
A Event studies around the adoption of Say on Pay regulations
As mentioned in the Introduction on April 20 2007 the House of Representatives passed a SOP
Bill by a 2-1 margin On the same day then-Senator Barack Obama introduced a companion
bill (S1181) in the Senate (which was then put on hold by the Senate Banking Committee)
While the SOP Billrsquos approval was expected (Democrats were in control of the House and
supported the SOP Bill) the 2-1 margin was unexpected suggesting some support for SOP
among Republicans
Cai and Walkling (2011) examine the market reaction to the Housersquos approval of the SOP Bill
for a sample of firms in the SampP 1500 index and document a positive reaction for firms more
likely to benefit from greater shareholder voice over executive pay namely firms with high
abnormal CEO cash pay (defined as the difference between actual CEO cash pay and the level
predicted based on known economic determinants such a size performance industry) firms
with low pay-for-performance sensitivity firms with a history of shareholders willing to vote
against compensation-related management proposals and firms with a history of responsiveness
to shareholder pressure on compensation issues In other words they find a positive reaction in
firms where the compensation problems are more severe and a say on pay vote is more likely to
trigger a response (eg more shareholders willing to vote against management and greater firmsrsquo
propensity to respond to an adverse vote) Note however that they do not find a significant
impact for firms with high abnormal equity and total CEO pay
Larcker Ormazabal and Taylor (2011) also examine the market reaction to the Housersquos approval
of the SOP Bill (in addition to other regulatory events related to executive pay and other
governance provisions) Similar to Cai and Wakling (2011) they fail to find a significant impact
for firms with high abnormal total CEO pay (they do not examine the price reaction for firms
with abnormal cash CEO pay or firms with lower pay-performance sensitivity)
A concern with both studies is that the Housersquos approval of a bill does not guarantee passage in
the Senate and approval by the White House In fact the press at the time reported that the
prospects of the bill in the Senate were uncertain (New York Times 2007) and the Bush White
House openly opposed the Bill (Associated Press 2007) Hence it is not clear the extent to
which this event increased the likelihood of a say on pay legislation Arguably as noted in
Section I the only lsquoeventrsquo that substantially increased the likelihood of say on pay legislation
was the financial crisis and the related perception of compensation abuses (eg outrage over AIG
bonuses)
Ferri and Maber (2013) argue that the announcement of the submission of SOP regulation to the
Parliament in the United Kingdom in June 2002 offers a more powerful setting for an event
study since it was largely unexpected and increased substantially the probability of SOP
adoption (Parliamentrsquos approval was virtually guaranteed) Using this event they document
positive abnormal returns for firms with excess CEO pay combined with poor performance and
for firms with controversial pay practices that weaken the penalties for poor performance (those
practices were often removed in response to SOP votes as discussed earlier in Section IIB)
They interpret their findings as consistent with shareholders perceiving SOP as a value
enhancing monitoring mechanism for firms with low pay-to-poor-performance sensitivity
Finally a recent study by Iliev and Vitanova (2013) examines the market reaction around the
announcement of the SEC final SOP rule that on January 25 2011 exempted small firms (ie
firms with a public float below $75 million) from adopting SOP for two years (press reports at
the time noted that the exemption could become permanent) An appealing feature of this setting
is that the SEC decision was an unexpected reversal from the rule proposed in October 2010 rule
under which no publicly traded firm would be exempted
Iliev and Vitanova (2013) compare the market reaction to this announcement for firms just above
the SEC-imposed threshold and thus subject to the SOP rule and firms just below the threshold
(but otherwise quite similar) and thus exempted from SOP While they find not significant
abnormal returns for either group the abnormal returns for exempted firms are significantly
more negative a finding that the authors interpret as evidence of a positive value effect of
mandatory SOP While the setting is an interesting one some concerns remain as exempted
firms are fairly small and the magnitude of the executive pay problem (if any) is unlikely to be
large enough to explain the differential market reaction Also the same study finds no effect on
CEO pay level and composition (see Section IIA) Hence the reason for the differential stock
price reaction remains unclear It would be helpful to know if the event study result is driven by
(or stronger for) exempted firms with excessive pay or problematic pay practices
B Event studies around Shareholder Proposals to Adopt Say on Pay
Starting in 2006 shareholder activists led by union pension funds submitted shareholder
proposal to adopt SOP at hundreds of US firms in an attempt to induce voluntary or mandatory
widespread adoption of SOP
Cai and Walkling (2011) examine the market reaction to proxy filings and annual meetings of
113 firms targeted by nonbinding shareholder proposals to adopt SOP between 2006 and 2008
On average they find insignificant returns around both the submission of the proposal (proxy
filing date) and the vote on the proposal (annual meeting date) But they also find that when the
proposals are filed by union pension funds the abnormal returns (still insignificant) are more
negative than when filed by other activists and that when the proposal is defeated the abnormal
returns (while insignificant) are more positive than when the proposal is passed (it is not clear
whether this result is driven by union-sponsored proposals) They interpret these findings as
evidence that the market views SOP proposals filed by union pension funds are driven by special
interests rather than value maximization10
However caution is required in interpreting this
evidence Virtually all activists filing SOP proposals were coordinated by a group of investors
(mostly but not only union pension funds) who made available to other activists a template for
SOP proposals and a list of target firms (Ferri and Weber 2009) Hence the distinction between
union and non-union proponents in the context of SOP proposals may be only apparent More
importantly the list of target firms was publicly available months before the proxy filing dates
so it is not clear whether the proxy statements contained any new information Similarly the
voting outcome of the SOP proposals may largely be anticipated (based on the composition of
institutional owners proxy advisorsrsquo recommendations etc) Finally the analyses do not control
for other (potentially new) information contained in proxy statements (eg executive
compensation report other items up for a vote at the annual meeting) or other events occurring at
the annual meeting (eg shareholder votes on other items)
Cuntildeat Gine and Guadalupe (2013) also examine the market reaction to the voting outcome of
nonbinding shareholder proposals to adopt SOP but to alleviate these concerns they employ a
lsquofuzzyrsquo regression discontinuity design (RDD) essentially comparing the stock price reaction to
SOP proposals that pass by a small margin to the reaction to SOP proposals that fail by a small
10
Consistent with this interpretation the authors also find that SOP proposals generally targeted larger firms rather
than firms with excessive pay or lower pay-performance sensitivity However as noted by Ferri and Sandino (2009)
activists typically submit shareholder proposals aimed at promoting policy reforms at a broad sample of large firms
rather than the firms that would most benefit from the proposals because they believe that showing widespread
support for the proposal across all types of firms enhances its credibility with policy makers
margin (Cuntildeat Gine and Guadalupe 2012 Ertimur Ferri and Oesch forthcoming) The underlying idea
is that firms around the threshold are likely to have similar characteristics (indeed they do as the
authors show) but differ in the likelihood of implementation which is typically much higher for
proposals passing the threshold (Ertimur Ferri and Stubben 2010 Thomas and Cotter 2007)
Indeed the authors show that the likelihood of subsequently adopting SOP is 48 higher when
the SOP proposal passes by a small margin relatively to when it fails by a small margin Besides
because in these close-call situations the voting outcome is uncertain its resolution (pass or fail)
is likely to convey new information about the likelihood of SOP adoption making this setting
suitable to an event study11
Using this RDD approach Cuntildeat Gine and Guadalupe (2013) find that on the day of the vote a
SOP proposal that passes by a small margin yields an abnormal return of 24 relative to one
that fails (after controlling for other proposals voted upon at the same meeting) and estimate the
lsquofullrsquo value of SOP at 46 (after taking into account the increase in the probability of SOP
implementation)12
Aside from the issues discussed in Section IA (ie generalizability to other
firms) this estimate seems too large to reflect the present value of future reductions in excess
CEO pay (as noted by the authors their estimate of the value of SOP is equivalent to the
estimated value of removing two anti-takeover provision based on Cuntildeat Gine and Guadalupe
2012) particularly because (i) the authors find no evidence of subsequent changes in levels and
composition of CEO pay for firms implementing SOP (see Section IIA) an (ii) the sample firms
11
Consistent with this observation the authors find no market reaction to the voting outcome of SOP shareholder
proposals away from the threshold
12 Since the outcome of the vote is not binding the 24 market reaction only reflects the expected increase in the
probability of SOP adoption and thus understates the value of the SOP provision
do not appear to be characterized by excess CEO pay However the authors also find significant
improvements in subsequent operating performance and efficiency as a result of passing SOP
proposals and thus conjecture that the SOP vote is viewed as a tool to express a vote of
confidence in management performance more than a way to change compensation practices and
that the large positive market reaction reflects expected performance improvements under this
tighter monitoring regime While this is an interesting idea it remains unclear what additional
ldquoteethrdquo a SOP vote may provide over existing (non-compensation related) monitoring
mechanisms shareholders can express (and do often express) their dissatisfaction with
management performance when voting on director elections (Cai Garner and Walkling 2009)
C Event studies around compensation changes induced by SOP
Two studies examine announcements of changes to compensation plans related to SOP Larcker
McCall and Ormazabal (2013) examine the market reaction to (non-contaminated) compensation
changes disclosed in 8-K filings by Russell 3000 firms in the US during the year before the first
SOP vote They find that changes that appear to be made to avoid a negative proxy advisorrsquos
recommendation and thus a negative SOP vote are associated with a negative abnormal return
of -044 whereas other compensation changes are not associated with a significant stock price
reaction Ertimur Ferri and Oesch (2013) perform a similar analysis but focusing on
compensation changes made after the first SOP vote (and explicitly in response to the vote) by
firms receiving a negative recommendation and a high voting dissent in 2011 They fail to find
significant abnormal returns even for the subset of compensation changes that resulted in a
positive recommendation and low dissent in 2012 (and thus were presumably perceived to be
adequate and material by proxy advisors and voting shareholders)
D Other approaches effect of SOP on Tobinrsquos Q
In their cross-country study Correa and Lel (2013) also examine the effect of SOP laws on
Tobinrsquos Q and report a 36 increase in firm value following the adoption of the SOP laws
(more precisely a 36 higher firm value for post-SOP observations relative to a control sample
that pools together pre-SOP and non-SOP observations) They acknowledge that this increase in
firm value is too large to be justified by the relative decrease in CEO pay that they document and
suggest (but do not test) that it may reflect better alignment of pay and performance13
A key
concern is that higher valuation in countries with SOP laws may reflect other governance
changes introduced at the same time While the study controls for other compensation-related
laws there may be other non-compensation regulations introduced with SOP and with a
potentially larger impact on firm value
IV What have we learned
Adoption of SOP in the US has been long advocated by those who contend that executive pay is
a manifestation of rather than a solution to the agency problem and by those who favor great
shareholder involvement in corporate decisions A growing body of research is starting to
investigate the effect of SOP on executive pay and firm value It is premature to draw definitive
conclusions also because these studies differ in methodologies and settings but three points
seem to emerge First there is robust evidence that boards respond to SOP votes when
13
They also find that the firm value increase is higher when the relative decrease in CEO pay results in a lower pay
differential between CEO and other top managers (CEO pay slice) implying that a source of value creation may be
the reduced pay inequality among the top management team But this seems unlikely to explain the change in
Tobinrsquos Q since the increase occurs for both firms with advisory SOP laws and firms with binding SOP laws and
there is no change in CEO pay slice in firms subject to binding SOP laws
shareholders use it ldquovoicerdquo is heardrdquo Both in the US and UK studies show that not only firms
failing to win the SOP vote but also firms facing substantial dissent (20-30 of the votes
against) make changes to their compensation plans in consultation with institutional investors
and proxy advisors Second in both the US and UK institutional investors appear to use the
power of SOP votes to pressure firms into strengthening the link (or the perceived link) between
pay and performance (particularly on the downside) but not to pressure firms to reduce target
levels of pay suggesting a reluctance of institutional investors to lsquoregulatersquo executive pay Third
and consistent with the above most studies examining the aggregate effect of SOP on CEO pay
find some increase in pay-performance sensitivity but not effect on the level and growth of CEO
pay While some observers may interpret the lack of an effect on CEO pay levels as evidence of
the failure of SOP one should note that SOP per se is a neutral tool Its impact depends on what
investors choose to lsquosay on payrsquo Fourth based on my interpretation of the existing research to
date there is little evidence of an effect of SOP on firm value The studies documenting a
positive price impact appear subject to alternative interpretations or not plausible in their
findings (eg they fail to find an effect on CEO pay or the effect is too small to justify the
documented price impact hence it remains unclear what the source of value creation is) Also
there seems to be no stock price reaction to the SOP-induced compensation changes Finally
given the nature of these SOP-induced changes while some of them may be beneficial it seems
hard to believe that they will have a statistically detectable impact on firm value except perhaps
in a handful of firms where the compensation plan is subject to a complete overhaul
Perhaps one explanation for the weak impact of SOP is that executive pay problems were
overstated or that by the time SOP was introduced (more than a decade after the Enron-type
scandals that led to calls for greater shareholder voice) they had been already addressed via
other mechanisms (hedge fund activism monitoring by institutional investors vote-no
campaigns against compensation committee members SEC-mandated pay disclosures)
Overall though based on the evidence to date it does not appear that SOP had a major impact14
Was the adoption of SOP an lsquooptimalrsquo choice from a social welfare point of view This is a
difficult question to answer given the challenges of identifying and measuring all the potential
costs and benefits associated with SOP (likely to change over time and differ across countries)
Perhaps the most positive view of SOP is that its introduction prevented the adoption of other
more radical and intrusive regulatory measures (eg CEO pay caps) In that sense it may be
viewed as an lsquooptimalrsquo answer (ie the lesser of two evils) to the political pressure to reform
executive pay during the financial crisis
14
This interpretation of the evidence is also consistent with the prediction of analytical studies Ozbas and
Matsusaka (2013) model the benefits and costs of shareholder empowerment distinguishing between the right to
approve and the right to propose They show that permitting shareholders to propose directors or policies can cause
value-maximizing managers to take value-reducing actions to accommodate activist investors with non-value-
maximizing goals As for the right to approve it is weakly beneficial that is approval rights (such as a SOP vote)
are generally beneficial for shareholders in that they limit the managerrsquos ability to pursue private benefits at
shareholder expense but they have minimal impact on managerial actions and firm value (because the manager in
effect can threaten shareholders with an undesirable status quo if they do not approve the managerrsquos proposed
action)
References
Associated Press 2007 Administration opposes say on pay bill April 19
Bainbridge S 2008 Remarks on say on pay An unjustified incursion on director authority
UCLA School of Law Research Paper No 08-06
Bebchuk L 2005 The Case for Increasing Shareholder Power Harvard Law Review 118833-
917
Bebchuk L 2007 Written testimony submitted before the Committee on Financial Services
United States House of Representatives Hearing on Empowering Shareholders on Executive
Compensation March 8
Bebchuk LA A Cohen and A Ferrell 2009 What Matters in Corporate Governance Review
of Financial Studies 22783-827
Cai J J Garner and R Walkling 2009 Electing Directors Journal of Finance 642387-2419
Cai J and R Walkling 2011 Shareholdersrsquo Say on Pay Does it Create Value Journal of
Financial and Quantitative Analysis 46299ndash339
Coates IV JC (2009) Testimony before the Subcommittee on Securities Insurance and
Investment of the Committee on Banking Housing and Urban Affairs United States Senate
July 29 2009 httppapersssrncomsol3paperscfmabstract_id=1475355
Correa R and U Lel 2013 Say On Pay Laws Executive Compensation CEO Pay Slice and
Firm Value Around the World Federal Reserve Board Working Paper
Cuntildeat V M Gine and M Guadalupe 2012 The Vote is Cast The Effect of Corporate
Governance on Shareholder Value Journal of Finance 671943-1977
Cuntildeat V M Gine and M Guadalupe 2013 Say Pays Shareholder Voice and Firm
Performance ECGI - Finance Working Paper No 373
Del Guercio D L Seery and T Woidtke 2008 Do Boards Pay Attention When Institutional
Investors lsquoJust Vote Norsquo Journal of Financial Economics 9084-103
Ertimur Y F Ferri and S Stubben 2010 Board of Directorslsquo Responsiveness to Shareholders
Evidence from Shareholder Proposals Journal of Corporate Finance 1653-72
Ertimur Y F Ferri and V Muslu 2011 Shareholder Activism and CEO Pay Review of
Financial Studies 24535ndash592
Ertimur Y F Ferri and D Oesch 2013 Shareholder Votes and Proxy Advisors Evidence from
Say on Pay Journal of Accounting Research 51951-996
Ertimur Y F Ferri and D Oesch forthcoming Does the Director Election System Matter
Evidence from Majority Voting Review of Accounting Studies forthcoming
Ferri F 2012 Low-Cost Shareholder Activism A Review of the Evidence Research
Handbook on the Economics of Corporate Law ed CA Hill and BH McDonnell
Edward Elgar Publishing
Ferri F and D Maber 2013 Say on Pay Votes and CEO Compensation Evidence from the UK
Review of Finance 17527-563
Ferri F and D Oesch 2013 Management Influence on Investors Evidence from Shareholder
Votes on the Frequency of Say on Pay Columbia Business School Research Paper No 13-17
Ferri F and T Sandino 2009 The impact of shareholder activism on financial reporting and
compensation The case of employee stock options expensing The Accounting Review 84433ndash
466
Ferri F and J Weber 2009 AFSCME vs Mozilohellipand Say on Pay for All (A) Harvard
Business School Case 109-009
Gompers P J Ishii and A Metrick 2003 Corporate Governance and Equity Prices Quarterly
Journal of Economics 118107-155
Gordon J 2009 Say on Paylsquo Cautionary Notes on the UK Experience and the Case for
Shareholders Opt-In Harvard Journal on Legislation 46323-368
Iliev P and S Vitanova 2013 The effect of the Say on Pay in the US Pennsylvania State
University Working Paper
Kaplan S 2007 Testimony of Steven N Kaplan on Empowering Shareholders on Executive
Compensation and HR 1257 the Shareholder Vote on Executive Compensation Act before the
Committee on Financial Services United States House of Representatives March 8
Larcker D A McCall and G Ormazabal 2013 The Economic Consequences of Proxy
Advisor Say-on-Pay Voting Policies Working Paper Stanford University
Larcker D G Ormazabal and D Taylor 2011 The market reaction to corporate governance
regulation Journal of Financial Economics 101431ndash448
Levit D and N Malenko 2011 Non-binding voting for shareholder proposals Journal of
Finance 661579ndash1614
Mishel L and N Sabadish 2012 How executive compensation and financial-sector pay have
fueled income inequality Issue Brief 331 Economic Policy Institute
Murphy K 2012 The Politics of Pay A Legislative History of Executive Compensation in
Jennifer Hill and Randall Thomas eds Research Handbook on Executive Pay Edward Elgar
Publishers
New York Times 2007 House votes to give investors say on executive pay April 21
Norris F 2004 Corporate democracy and the power to embarrass New York Times March 4
Ozbas O and J Matsusaka2013 Managerial Accommodation Proxy Access and the Cost of
Shareholder Empowerment University of Southern California CLEO Research Paper Series No
C12-1
Thomas R and J Cotter 2007 Shareholder proposals in the new millennium Shareholder
support board response and market reaction Journal of Corporate Finance 13 368ndash391
Thomas R A Palmiter and J Cotter 2012 Dodd-Franks Say on Pay Will it Lead to a Greater
Role for Shareholders in Corporate Governance Cornell Law Review 971213-1266
Thomas R and C Van der Elst 2013 The International Scope of Say on Pay ECGI Law
Working Paper No227
investors (iii) firms making changes to their compensation plans receive greater voting support
at the following SOP vote
Ertimur Ferri and Oesch (2013) also highlight the strong influence of the recommendations
released by proxy advisors (particularly Institutional Shareholder Services ISS) with a negative
recommendation being associated with 25 more votes against the compensation plan Further
evidence of the significant influence of ISS is the striking discontinuity in the relation between
the extent of SOP voting dissent and firmsrsquo responsiveness to the vote the percentage of firms
making compensation changes in response to SOP votes jumps from 32 to 72 around the
30 voting dissent threshold Why After the 2011 proxy season ISS had indicated that firms
failing to ldquoadequatelyrdquo respond to SOP voting dissent above 30 would receive a negative
recommendation in 2012 on the SOP proposal and on the election of compensation committee
members
Finally similar to the UK experience there is also anecdotal evidence of analogous
compensation changes being made by US firms ahead of the SOP vote to avoid a negative
proxy advisor recommendation and an adverse shareholder vote (Thomas Palmiter and Cotter
2013 Larcker McCall and Ormazabal 2013)
III The Effect of Say on Pay on Firm Value
In this section we examine the empirical studies on the overall effect of SOP on firm value
A Event studies around the adoption of Say on Pay regulations
As mentioned in the Introduction on April 20 2007 the House of Representatives passed a SOP
Bill by a 2-1 margin On the same day then-Senator Barack Obama introduced a companion
bill (S1181) in the Senate (which was then put on hold by the Senate Banking Committee)
While the SOP Billrsquos approval was expected (Democrats were in control of the House and
supported the SOP Bill) the 2-1 margin was unexpected suggesting some support for SOP
among Republicans
Cai and Walkling (2011) examine the market reaction to the Housersquos approval of the SOP Bill
for a sample of firms in the SampP 1500 index and document a positive reaction for firms more
likely to benefit from greater shareholder voice over executive pay namely firms with high
abnormal CEO cash pay (defined as the difference between actual CEO cash pay and the level
predicted based on known economic determinants such a size performance industry) firms
with low pay-for-performance sensitivity firms with a history of shareholders willing to vote
against compensation-related management proposals and firms with a history of responsiveness
to shareholder pressure on compensation issues In other words they find a positive reaction in
firms where the compensation problems are more severe and a say on pay vote is more likely to
trigger a response (eg more shareholders willing to vote against management and greater firmsrsquo
propensity to respond to an adverse vote) Note however that they do not find a significant
impact for firms with high abnormal equity and total CEO pay
Larcker Ormazabal and Taylor (2011) also examine the market reaction to the Housersquos approval
of the SOP Bill (in addition to other regulatory events related to executive pay and other
governance provisions) Similar to Cai and Wakling (2011) they fail to find a significant impact
for firms with high abnormal total CEO pay (they do not examine the price reaction for firms
with abnormal cash CEO pay or firms with lower pay-performance sensitivity)
A concern with both studies is that the Housersquos approval of a bill does not guarantee passage in
the Senate and approval by the White House In fact the press at the time reported that the
prospects of the bill in the Senate were uncertain (New York Times 2007) and the Bush White
House openly opposed the Bill (Associated Press 2007) Hence it is not clear the extent to
which this event increased the likelihood of a say on pay legislation Arguably as noted in
Section I the only lsquoeventrsquo that substantially increased the likelihood of say on pay legislation
was the financial crisis and the related perception of compensation abuses (eg outrage over AIG
bonuses)
Ferri and Maber (2013) argue that the announcement of the submission of SOP regulation to the
Parliament in the United Kingdom in June 2002 offers a more powerful setting for an event
study since it was largely unexpected and increased substantially the probability of SOP
adoption (Parliamentrsquos approval was virtually guaranteed) Using this event they document
positive abnormal returns for firms with excess CEO pay combined with poor performance and
for firms with controversial pay practices that weaken the penalties for poor performance (those
practices were often removed in response to SOP votes as discussed earlier in Section IIB)
They interpret their findings as consistent with shareholders perceiving SOP as a value
enhancing monitoring mechanism for firms with low pay-to-poor-performance sensitivity
Finally a recent study by Iliev and Vitanova (2013) examines the market reaction around the
announcement of the SEC final SOP rule that on January 25 2011 exempted small firms (ie
firms with a public float below $75 million) from adopting SOP for two years (press reports at
the time noted that the exemption could become permanent) An appealing feature of this setting
is that the SEC decision was an unexpected reversal from the rule proposed in October 2010 rule
under which no publicly traded firm would be exempted
Iliev and Vitanova (2013) compare the market reaction to this announcement for firms just above
the SEC-imposed threshold and thus subject to the SOP rule and firms just below the threshold
(but otherwise quite similar) and thus exempted from SOP While they find not significant
abnormal returns for either group the abnormal returns for exempted firms are significantly
more negative a finding that the authors interpret as evidence of a positive value effect of
mandatory SOP While the setting is an interesting one some concerns remain as exempted
firms are fairly small and the magnitude of the executive pay problem (if any) is unlikely to be
large enough to explain the differential market reaction Also the same study finds no effect on
CEO pay level and composition (see Section IIA) Hence the reason for the differential stock
price reaction remains unclear It would be helpful to know if the event study result is driven by
(or stronger for) exempted firms with excessive pay or problematic pay practices
B Event studies around Shareholder Proposals to Adopt Say on Pay
Starting in 2006 shareholder activists led by union pension funds submitted shareholder
proposal to adopt SOP at hundreds of US firms in an attempt to induce voluntary or mandatory
widespread adoption of SOP
Cai and Walkling (2011) examine the market reaction to proxy filings and annual meetings of
113 firms targeted by nonbinding shareholder proposals to adopt SOP between 2006 and 2008
On average they find insignificant returns around both the submission of the proposal (proxy
filing date) and the vote on the proposal (annual meeting date) But they also find that when the
proposals are filed by union pension funds the abnormal returns (still insignificant) are more
negative than when filed by other activists and that when the proposal is defeated the abnormal
returns (while insignificant) are more positive than when the proposal is passed (it is not clear
whether this result is driven by union-sponsored proposals) They interpret these findings as
evidence that the market views SOP proposals filed by union pension funds are driven by special
interests rather than value maximization10
However caution is required in interpreting this
evidence Virtually all activists filing SOP proposals were coordinated by a group of investors
(mostly but not only union pension funds) who made available to other activists a template for
SOP proposals and a list of target firms (Ferri and Weber 2009) Hence the distinction between
union and non-union proponents in the context of SOP proposals may be only apparent More
importantly the list of target firms was publicly available months before the proxy filing dates
so it is not clear whether the proxy statements contained any new information Similarly the
voting outcome of the SOP proposals may largely be anticipated (based on the composition of
institutional owners proxy advisorsrsquo recommendations etc) Finally the analyses do not control
for other (potentially new) information contained in proxy statements (eg executive
compensation report other items up for a vote at the annual meeting) or other events occurring at
the annual meeting (eg shareholder votes on other items)
Cuntildeat Gine and Guadalupe (2013) also examine the market reaction to the voting outcome of
nonbinding shareholder proposals to adopt SOP but to alleviate these concerns they employ a
lsquofuzzyrsquo regression discontinuity design (RDD) essentially comparing the stock price reaction to
SOP proposals that pass by a small margin to the reaction to SOP proposals that fail by a small
10
Consistent with this interpretation the authors also find that SOP proposals generally targeted larger firms rather
than firms with excessive pay or lower pay-performance sensitivity However as noted by Ferri and Sandino (2009)
activists typically submit shareholder proposals aimed at promoting policy reforms at a broad sample of large firms
rather than the firms that would most benefit from the proposals because they believe that showing widespread
support for the proposal across all types of firms enhances its credibility with policy makers
margin (Cuntildeat Gine and Guadalupe 2012 Ertimur Ferri and Oesch forthcoming) The underlying idea
is that firms around the threshold are likely to have similar characteristics (indeed they do as the
authors show) but differ in the likelihood of implementation which is typically much higher for
proposals passing the threshold (Ertimur Ferri and Stubben 2010 Thomas and Cotter 2007)
Indeed the authors show that the likelihood of subsequently adopting SOP is 48 higher when
the SOP proposal passes by a small margin relatively to when it fails by a small margin Besides
because in these close-call situations the voting outcome is uncertain its resolution (pass or fail)
is likely to convey new information about the likelihood of SOP adoption making this setting
suitable to an event study11
Using this RDD approach Cuntildeat Gine and Guadalupe (2013) find that on the day of the vote a
SOP proposal that passes by a small margin yields an abnormal return of 24 relative to one
that fails (after controlling for other proposals voted upon at the same meeting) and estimate the
lsquofullrsquo value of SOP at 46 (after taking into account the increase in the probability of SOP
implementation)12
Aside from the issues discussed in Section IA (ie generalizability to other
firms) this estimate seems too large to reflect the present value of future reductions in excess
CEO pay (as noted by the authors their estimate of the value of SOP is equivalent to the
estimated value of removing two anti-takeover provision based on Cuntildeat Gine and Guadalupe
2012) particularly because (i) the authors find no evidence of subsequent changes in levels and
composition of CEO pay for firms implementing SOP (see Section IIA) an (ii) the sample firms
11
Consistent with this observation the authors find no market reaction to the voting outcome of SOP shareholder
proposals away from the threshold
12 Since the outcome of the vote is not binding the 24 market reaction only reflects the expected increase in the
probability of SOP adoption and thus understates the value of the SOP provision
do not appear to be characterized by excess CEO pay However the authors also find significant
improvements in subsequent operating performance and efficiency as a result of passing SOP
proposals and thus conjecture that the SOP vote is viewed as a tool to express a vote of
confidence in management performance more than a way to change compensation practices and
that the large positive market reaction reflects expected performance improvements under this
tighter monitoring regime While this is an interesting idea it remains unclear what additional
ldquoteethrdquo a SOP vote may provide over existing (non-compensation related) monitoring
mechanisms shareholders can express (and do often express) their dissatisfaction with
management performance when voting on director elections (Cai Garner and Walkling 2009)
C Event studies around compensation changes induced by SOP
Two studies examine announcements of changes to compensation plans related to SOP Larcker
McCall and Ormazabal (2013) examine the market reaction to (non-contaminated) compensation
changes disclosed in 8-K filings by Russell 3000 firms in the US during the year before the first
SOP vote They find that changes that appear to be made to avoid a negative proxy advisorrsquos
recommendation and thus a negative SOP vote are associated with a negative abnormal return
of -044 whereas other compensation changes are not associated with a significant stock price
reaction Ertimur Ferri and Oesch (2013) perform a similar analysis but focusing on
compensation changes made after the first SOP vote (and explicitly in response to the vote) by
firms receiving a negative recommendation and a high voting dissent in 2011 They fail to find
significant abnormal returns even for the subset of compensation changes that resulted in a
positive recommendation and low dissent in 2012 (and thus were presumably perceived to be
adequate and material by proxy advisors and voting shareholders)
D Other approaches effect of SOP on Tobinrsquos Q
In their cross-country study Correa and Lel (2013) also examine the effect of SOP laws on
Tobinrsquos Q and report a 36 increase in firm value following the adoption of the SOP laws
(more precisely a 36 higher firm value for post-SOP observations relative to a control sample
that pools together pre-SOP and non-SOP observations) They acknowledge that this increase in
firm value is too large to be justified by the relative decrease in CEO pay that they document and
suggest (but do not test) that it may reflect better alignment of pay and performance13
A key
concern is that higher valuation in countries with SOP laws may reflect other governance
changes introduced at the same time While the study controls for other compensation-related
laws there may be other non-compensation regulations introduced with SOP and with a
potentially larger impact on firm value
IV What have we learned
Adoption of SOP in the US has been long advocated by those who contend that executive pay is
a manifestation of rather than a solution to the agency problem and by those who favor great
shareholder involvement in corporate decisions A growing body of research is starting to
investigate the effect of SOP on executive pay and firm value It is premature to draw definitive
conclusions also because these studies differ in methodologies and settings but three points
seem to emerge First there is robust evidence that boards respond to SOP votes when
13
They also find that the firm value increase is higher when the relative decrease in CEO pay results in a lower pay
differential between CEO and other top managers (CEO pay slice) implying that a source of value creation may be
the reduced pay inequality among the top management team But this seems unlikely to explain the change in
Tobinrsquos Q since the increase occurs for both firms with advisory SOP laws and firms with binding SOP laws and
there is no change in CEO pay slice in firms subject to binding SOP laws
shareholders use it ldquovoicerdquo is heardrdquo Both in the US and UK studies show that not only firms
failing to win the SOP vote but also firms facing substantial dissent (20-30 of the votes
against) make changes to their compensation plans in consultation with institutional investors
and proxy advisors Second in both the US and UK institutional investors appear to use the
power of SOP votes to pressure firms into strengthening the link (or the perceived link) between
pay and performance (particularly on the downside) but not to pressure firms to reduce target
levels of pay suggesting a reluctance of institutional investors to lsquoregulatersquo executive pay Third
and consistent with the above most studies examining the aggregate effect of SOP on CEO pay
find some increase in pay-performance sensitivity but not effect on the level and growth of CEO
pay While some observers may interpret the lack of an effect on CEO pay levels as evidence of
the failure of SOP one should note that SOP per se is a neutral tool Its impact depends on what
investors choose to lsquosay on payrsquo Fourth based on my interpretation of the existing research to
date there is little evidence of an effect of SOP on firm value The studies documenting a
positive price impact appear subject to alternative interpretations or not plausible in their
findings (eg they fail to find an effect on CEO pay or the effect is too small to justify the
documented price impact hence it remains unclear what the source of value creation is) Also
there seems to be no stock price reaction to the SOP-induced compensation changes Finally
given the nature of these SOP-induced changes while some of them may be beneficial it seems
hard to believe that they will have a statistically detectable impact on firm value except perhaps
in a handful of firms where the compensation plan is subject to a complete overhaul
Perhaps one explanation for the weak impact of SOP is that executive pay problems were
overstated or that by the time SOP was introduced (more than a decade after the Enron-type
scandals that led to calls for greater shareholder voice) they had been already addressed via
other mechanisms (hedge fund activism monitoring by institutional investors vote-no
campaigns against compensation committee members SEC-mandated pay disclosures)
Overall though based on the evidence to date it does not appear that SOP had a major impact14
Was the adoption of SOP an lsquooptimalrsquo choice from a social welfare point of view This is a
difficult question to answer given the challenges of identifying and measuring all the potential
costs and benefits associated with SOP (likely to change over time and differ across countries)
Perhaps the most positive view of SOP is that its introduction prevented the adoption of other
more radical and intrusive regulatory measures (eg CEO pay caps) In that sense it may be
viewed as an lsquooptimalrsquo answer (ie the lesser of two evils) to the political pressure to reform
executive pay during the financial crisis
14
This interpretation of the evidence is also consistent with the prediction of analytical studies Ozbas and
Matsusaka (2013) model the benefits and costs of shareholder empowerment distinguishing between the right to
approve and the right to propose They show that permitting shareholders to propose directors or policies can cause
value-maximizing managers to take value-reducing actions to accommodate activist investors with non-value-
maximizing goals As for the right to approve it is weakly beneficial that is approval rights (such as a SOP vote)
are generally beneficial for shareholders in that they limit the managerrsquos ability to pursue private benefits at
shareholder expense but they have minimal impact on managerial actions and firm value (because the manager in
effect can threaten shareholders with an undesirable status quo if they do not approve the managerrsquos proposed
action)
References
Associated Press 2007 Administration opposes say on pay bill April 19
Bainbridge S 2008 Remarks on say on pay An unjustified incursion on director authority
UCLA School of Law Research Paper No 08-06
Bebchuk L 2005 The Case for Increasing Shareholder Power Harvard Law Review 118833-
917
Bebchuk L 2007 Written testimony submitted before the Committee on Financial Services
United States House of Representatives Hearing on Empowering Shareholders on Executive
Compensation March 8
Bebchuk LA A Cohen and A Ferrell 2009 What Matters in Corporate Governance Review
of Financial Studies 22783-827
Cai J J Garner and R Walkling 2009 Electing Directors Journal of Finance 642387-2419
Cai J and R Walkling 2011 Shareholdersrsquo Say on Pay Does it Create Value Journal of
Financial and Quantitative Analysis 46299ndash339
Coates IV JC (2009) Testimony before the Subcommittee on Securities Insurance and
Investment of the Committee on Banking Housing and Urban Affairs United States Senate
July 29 2009 httppapersssrncomsol3paperscfmabstract_id=1475355
Correa R and U Lel 2013 Say On Pay Laws Executive Compensation CEO Pay Slice and
Firm Value Around the World Federal Reserve Board Working Paper
Cuntildeat V M Gine and M Guadalupe 2012 The Vote is Cast The Effect of Corporate
Governance on Shareholder Value Journal of Finance 671943-1977
Cuntildeat V M Gine and M Guadalupe 2013 Say Pays Shareholder Voice and Firm
Performance ECGI - Finance Working Paper No 373
Del Guercio D L Seery and T Woidtke 2008 Do Boards Pay Attention When Institutional
Investors lsquoJust Vote Norsquo Journal of Financial Economics 9084-103
Ertimur Y F Ferri and S Stubben 2010 Board of Directorslsquo Responsiveness to Shareholders
Evidence from Shareholder Proposals Journal of Corporate Finance 1653-72
Ertimur Y F Ferri and V Muslu 2011 Shareholder Activism and CEO Pay Review of
Financial Studies 24535ndash592
Ertimur Y F Ferri and D Oesch 2013 Shareholder Votes and Proxy Advisors Evidence from
Say on Pay Journal of Accounting Research 51951-996
Ertimur Y F Ferri and D Oesch forthcoming Does the Director Election System Matter
Evidence from Majority Voting Review of Accounting Studies forthcoming
Ferri F 2012 Low-Cost Shareholder Activism A Review of the Evidence Research
Handbook on the Economics of Corporate Law ed CA Hill and BH McDonnell
Edward Elgar Publishing
Ferri F and D Maber 2013 Say on Pay Votes and CEO Compensation Evidence from the UK
Review of Finance 17527-563
Ferri F and D Oesch 2013 Management Influence on Investors Evidence from Shareholder
Votes on the Frequency of Say on Pay Columbia Business School Research Paper No 13-17
Ferri F and T Sandino 2009 The impact of shareholder activism on financial reporting and
compensation The case of employee stock options expensing The Accounting Review 84433ndash
466
Ferri F and J Weber 2009 AFSCME vs Mozilohellipand Say on Pay for All (A) Harvard
Business School Case 109-009
Gompers P J Ishii and A Metrick 2003 Corporate Governance and Equity Prices Quarterly
Journal of Economics 118107-155
Gordon J 2009 Say on Paylsquo Cautionary Notes on the UK Experience and the Case for
Shareholders Opt-In Harvard Journal on Legislation 46323-368
Iliev P and S Vitanova 2013 The effect of the Say on Pay in the US Pennsylvania State
University Working Paper
Kaplan S 2007 Testimony of Steven N Kaplan on Empowering Shareholders on Executive
Compensation and HR 1257 the Shareholder Vote on Executive Compensation Act before the
Committee on Financial Services United States House of Representatives March 8
Larcker D A McCall and G Ormazabal 2013 The Economic Consequences of Proxy
Advisor Say-on-Pay Voting Policies Working Paper Stanford University
Larcker D G Ormazabal and D Taylor 2011 The market reaction to corporate governance
regulation Journal of Financial Economics 101431ndash448
Levit D and N Malenko 2011 Non-binding voting for shareholder proposals Journal of
Finance 661579ndash1614
Mishel L and N Sabadish 2012 How executive compensation and financial-sector pay have
fueled income inequality Issue Brief 331 Economic Policy Institute
Murphy K 2012 The Politics of Pay A Legislative History of Executive Compensation in
Jennifer Hill and Randall Thomas eds Research Handbook on Executive Pay Edward Elgar
Publishers
New York Times 2007 House votes to give investors say on executive pay April 21
Norris F 2004 Corporate democracy and the power to embarrass New York Times March 4
Ozbas O and J Matsusaka2013 Managerial Accommodation Proxy Access and the Cost of
Shareholder Empowerment University of Southern California CLEO Research Paper Series No
C12-1
Thomas R and J Cotter 2007 Shareholder proposals in the new millennium Shareholder
support board response and market reaction Journal of Corporate Finance 13 368ndash391
Thomas R A Palmiter and J Cotter 2012 Dodd-Franks Say on Pay Will it Lead to a Greater
Role for Shareholders in Corporate Governance Cornell Law Review 971213-1266
Thomas R and C Van der Elst 2013 The International Scope of Say on Pay ECGI Law
Working Paper No227
While the SOP Billrsquos approval was expected (Democrats were in control of the House and
supported the SOP Bill) the 2-1 margin was unexpected suggesting some support for SOP
among Republicans
Cai and Walkling (2011) examine the market reaction to the Housersquos approval of the SOP Bill
for a sample of firms in the SampP 1500 index and document a positive reaction for firms more
likely to benefit from greater shareholder voice over executive pay namely firms with high
abnormal CEO cash pay (defined as the difference between actual CEO cash pay and the level
predicted based on known economic determinants such a size performance industry) firms
with low pay-for-performance sensitivity firms with a history of shareholders willing to vote
against compensation-related management proposals and firms with a history of responsiveness
to shareholder pressure on compensation issues In other words they find a positive reaction in
firms where the compensation problems are more severe and a say on pay vote is more likely to
trigger a response (eg more shareholders willing to vote against management and greater firmsrsquo
propensity to respond to an adverse vote) Note however that they do not find a significant
impact for firms with high abnormal equity and total CEO pay
Larcker Ormazabal and Taylor (2011) also examine the market reaction to the Housersquos approval
of the SOP Bill (in addition to other regulatory events related to executive pay and other
governance provisions) Similar to Cai and Wakling (2011) they fail to find a significant impact
for firms with high abnormal total CEO pay (they do not examine the price reaction for firms
with abnormal cash CEO pay or firms with lower pay-performance sensitivity)
A concern with both studies is that the Housersquos approval of a bill does not guarantee passage in
the Senate and approval by the White House In fact the press at the time reported that the
prospects of the bill in the Senate were uncertain (New York Times 2007) and the Bush White
House openly opposed the Bill (Associated Press 2007) Hence it is not clear the extent to
which this event increased the likelihood of a say on pay legislation Arguably as noted in
Section I the only lsquoeventrsquo that substantially increased the likelihood of say on pay legislation
was the financial crisis and the related perception of compensation abuses (eg outrage over AIG
bonuses)
Ferri and Maber (2013) argue that the announcement of the submission of SOP regulation to the
Parliament in the United Kingdom in June 2002 offers a more powerful setting for an event
study since it was largely unexpected and increased substantially the probability of SOP
adoption (Parliamentrsquos approval was virtually guaranteed) Using this event they document
positive abnormal returns for firms with excess CEO pay combined with poor performance and
for firms with controversial pay practices that weaken the penalties for poor performance (those
practices were often removed in response to SOP votes as discussed earlier in Section IIB)
They interpret their findings as consistent with shareholders perceiving SOP as a value
enhancing monitoring mechanism for firms with low pay-to-poor-performance sensitivity
Finally a recent study by Iliev and Vitanova (2013) examines the market reaction around the
announcement of the SEC final SOP rule that on January 25 2011 exempted small firms (ie
firms with a public float below $75 million) from adopting SOP for two years (press reports at
the time noted that the exemption could become permanent) An appealing feature of this setting
is that the SEC decision was an unexpected reversal from the rule proposed in October 2010 rule
under which no publicly traded firm would be exempted
Iliev and Vitanova (2013) compare the market reaction to this announcement for firms just above
the SEC-imposed threshold and thus subject to the SOP rule and firms just below the threshold
(but otherwise quite similar) and thus exempted from SOP While they find not significant
abnormal returns for either group the abnormal returns for exempted firms are significantly
more negative a finding that the authors interpret as evidence of a positive value effect of
mandatory SOP While the setting is an interesting one some concerns remain as exempted
firms are fairly small and the magnitude of the executive pay problem (if any) is unlikely to be
large enough to explain the differential market reaction Also the same study finds no effect on
CEO pay level and composition (see Section IIA) Hence the reason for the differential stock
price reaction remains unclear It would be helpful to know if the event study result is driven by
(or stronger for) exempted firms with excessive pay or problematic pay practices
B Event studies around Shareholder Proposals to Adopt Say on Pay
Starting in 2006 shareholder activists led by union pension funds submitted shareholder
proposal to adopt SOP at hundreds of US firms in an attempt to induce voluntary or mandatory
widespread adoption of SOP
Cai and Walkling (2011) examine the market reaction to proxy filings and annual meetings of
113 firms targeted by nonbinding shareholder proposals to adopt SOP between 2006 and 2008
On average they find insignificant returns around both the submission of the proposal (proxy
filing date) and the vote on the proposal (annual meeting date) But they also find that when the
proposals are filed by union pension funds the abnormal returns (still insignificant) are more
negative than when filed by other activists and that when the proposal is defeated the abnormal
returns (while insignificant) are more positive than when the proposal is passed (it is not clear
whether this result is driven by union-sponsored proposals) They interpret these findings as
evidence that the market views SOP proposals filed by union pension funds are driven by special
interests rather than value maximization10
However caution is required in interpreting this
evidence Virtually all activists filing SOP proposals were coordinated by a group of investors
(mostly but not only union pension funds) who made available to other activists a template for
SOP proposals and a list of target firms (Ferri and Weber 2009) Hence the distinction between
union and non-union proponents in the context of SOP proposals may be only apparent More
importantly the list of target firms was publicly available months before the proxy filing dates
so it is not clear whether the proxy statements contained any new information Similarly the
voting outcome of the SOP proposals may largely be anticipated (based on the composition of
institutional owners proxy advisorsrsquo recommendations etc) Finally the analyses do not control
for other (potentially new) information contained in proxy statements (eg executive
compensation report other items up for a vote at the annual meeting) or other events occurring at
the annual meeting (eg shareholder votes on other items)
Cuntildeat Gine and Guadalupe (2013) also examine the market reaction to the voting outcome of
nonbinding shareholder proposals to adopt SOP but to alleviate these concerns they employ a
lsquofuzzyrsquo regression discontinuity design (RDD) essentially comparing the stock price reaction to
SOP proposals that pass by a small margin to the reaction to SOP proposals that fail by a small
10
Consistent with this interpretation the authors also find that SOP proposals generally targeted larger firms rather
than firms with excessive pay or lower pay-performance sensitivity However as noted by Ferri and Sandino (2009)
activists typically submit shareholder proposals aimed at promoting policy reforms at a broad sample of large firms
rather than the firms that would most benefit from the proposals because they believe that showing widespread
support for the proposal across all types of firms enhances its credibility with policy makers
margin (Cuntildeat Gine and Guadalupe 2012 Ertimur Ferri and Oesch forthcoming) The underlying idea
is that firms around the threshold are likely to have similar characteristics (indeed they do as the
authors show) but differ in the likelihood of implementation which is typically much higher for
proposals passing the threshold (Ertimur Ferri and Stubben 2010 Thomas and Cotter 2007)
Indeed the authors show that the likelihood of subsequently adopting SOP is 48 higher when
the SOP proposal passes by a small margin relatively to when it fails by a small margin Besides
because in these close-call situations the voting outcome is uncertain its resolution (pass or fail)
is likely to convey new information about the likelihood of SOP adoption making this setting
suitable to an event study11
Using this RDD approach Cuntildeat Gine and Guadalupe (2013) find that on the day of the vote a
SOP proposal that passes by a small margin yields an abnormal return of 24 relative to one
that fails (after controlling for other proposals voted upon at the same meeting) and estimate the
lsquofullrsquo value of SOP at 46 (after taking into account the increase in the probability of SOP
implementation)12
Aside from the issues discussed in Section IA (ie generalizability to other
firms) this estimate seems too large to reflect the present value of future reductions in excess
CEO pay (as noted by the authors their estimate of the value of SOP is equivalent to the
estimated value of removing two anti-takeover provision based on Cuntildeat Gine and Guadalupe
2012) particularly because (i) the authors find no evidence of subsequent changes in levels and
composition of CEO pay for firms implementing SOP (see Section IIA) an (ii) the sample firms
11
Consistent with this observation the authors find no market reaction to the voting outcome of SOP shareholder
proposals away from the threshold
12 Since the outcome of the vote is not binding the 24 market reaction only reflects the expected increase in the
probability of SOP adoption and thus understates the value of the SOP provision
do not appear to be characterized by excess CEO pay However the authors also find significant
improvements in subsequent operating performance and efficiency as a result of passing SOP
proposals and thus conjecture that the SOP vote is viewed as a tool to express a vote of
confidence in management performance more than a way to change compensation practices and
that the large positive market reaction reflects expected performance improvements under this
tighter monitoring regime While this is an interesting idea it remains unclear what additional
ldquoteethrdquo a SOP vote may provide over existing (non-compensation related) monitoring
mechanisms shareholders can express (and do often express) their dissatisfaction with
management performance when voting on director elections (Cai Garner and Walkling 2009)
C Event studies around compensation changes induced by SOP
Two studies examine announcements of changes to compensation plans related to SOP Larcker
McCall and Ormazabal (2013) examine the market reaction to (non-contaminated) compensation
changes disclosed in 8-K filings by Russell 3000 firms in the US during the year before the first
SOP vote They find that changes that appear to be made to avoid a negative proxy advisorrsquos
recommendation and thus a negative SOP vote are associated with a negative abnormal return
of -044 whereas other compensation changes are not associated with a significant stock price
reaction Ertimur Ferri and Oesch (2013) perform a similar analysis but focusing on
compensation changes made after the first SOP vote (and explicitly in response to the vote) by
firms receiving a negative recommendation and a high voting dissent in 2011 They fail to find
significant abnormal returns even for the subset of compensation changes that resulted in a
positive recommendation and low dissent in 2012 (and thus were presumably perceived to be
adequate and material by proxy advisors and voting shareholders)
D Other approaches effect of SOP on Tobinrsquos Q
In their cross-country study Correa and Lel (2013) also examine the effect of SOP laws on
Tobinrsquos Q and report a 36 increase in firm value following the adoption of the SOP laws
(more precisely a 36 higher firm value for post-SOP observations relative to a control sample
that pools together pre-SOP and non-SOP observations) They acknowledge that this increase in
firm value is too large to be justified by the relative decrease in CEO pay that they document and
suggest (but do not test) that it may reflect better alignment of pay and performance13
A key
concern is that higher valuation in countries with SOP laws may reflect other governance
changes introduced at the same time While the study controls for other compensation-related
laws there may be other non-compensation regulations introduced with SOP and with a
potentially larger impact on firm value
IV What have we learned
Adoption of SOP in the US has been long advocated by those who contend that executive pay is
a manifestation of rather than a solution to the agency problem and by those who favor great
shareholder involvement in corporate decisions A growing body of research is starting to
investigate the effect of SOP on executive pay and firm value It is premature to draw definitive
conclusions also because these studies differ in methodologies and settings but three points
seem to emerge First there is robust evidence that boards respond to SOP votes when
13
They also find that the firm value increase is higher when the relative decrease in CEO pay results in a lower pay
differential between CEO and other top managers (CEO pay slice) implying that a source of value creation may be
the reduced pay inequality among the top management team But this seems unlikely to explain the change in
Tobinrsquos Q since the increase occurs for both firms with advisory SOP laws and firms with binding SOP laws and
there is no change in CEO pay slice in firms subject to binding SOP laws
shareholders use it ldquovoicerdquo is heardrdquo Both in the US and UK studies show that not only firms
failing to win the SOP vote but also firms facing substantial dissent (20-30 of the votes
against) make changes to their compensation plans in consultation with institutional investors
and proxy advisors Second in both the US and UK institutional investors appear to use the
power of SOP votes to pressure firms into strengthening the link (or the perceived link) between
pay and performance (particularly on the downside) but not to pressure firms to reduce target
levels of pay suggesting a reluctance of institutional investors to lsquoregulatersquo executive pay Third
and consistent with the above most studies examining the aggregate effect of SOP on CEO pay
find some increase in pay-performance sensitivity but not effect on the level and growth of CEO
pay While some observers may interpret the lack of an effect on CEO pay levels as evidence of
the failure of SOP one should note that SOP per se is a neutral tool Its impact depends on what
investors choose to lsquosay on payrsquo Fourth based on my interpretation of the existing research to
date there is little evidence of an effect of SOP on firm value The studies documenting a
positive price impact appear subject to alternative interpretations or not plausible in their
findings (eg they fail to find an effect on CEO pay or the effect is too small to justify the
documented price impact hence it remains unclear what the source of value creation is) Also
there seems to be no stock price reaction to the SOP-induced compensation changes Finally
given the nature of these SOP-induced changes while some of them may be beneficial it seems
hard to believe that they will have a statistically detectable impact on firm value except perhaps
in a handful of firms where the compensation plan is subject to a complete overhaul
Perhaps one explanation for the weak impact of SOP is that executive pay problems were
overstated or that by the time SOP was introduced (more than a decade after the Enron-type
scandals that led to calls for greater shareholder voice) they had been already addressed via
other mechanisms (hedge fund activism monitoring by institutional investors vote-no
campaigns against compensation committee members SEC-mandated pay disclosures)
Overall though based on the evidence to date it does not appear that SOP had a major impact14
Was the adoption of SOP an lsquooptimalrsquo choice from a social welfare point of view This is a
difficult question to answer given the challenges of identifying and measuring all the potential
costs and benefits associated with SOP (likely to change over time and differ across countries)
Perhaps the most positive view of SOP is that its introduction prevented the adoption of other
more radical and intrusive regulatory measures (eg CEO pay caps) In that sense it may be
viewed as an lsquooptimalrsquo answer (ie the lesser of two evils) to the political pressure to reform
executive pay during the financial crisis
14
This interpretation of the evidence is also consistent with the prediction of analytical studies Ozbas and
Matsusaka (2013) model the benefits and costs of shareholder empowerment distinguishing between the right to
approve and the right to propose They show that permitting shareholders to propose directors or policies can cause
value-maximizing managers to take value-reducing actions to accommodate activist investors with non-value-
maximizing goals As for the right to approve it is weakly beneficial that is approval rights (such as a SOP vote)
are generally beneficial for shareholders in that they limit the managerrsquos ability to pursue private benefits at
shareholder expense but they have minimal impact on managerial actions and firm value (because the manager in
effect can threaten shareholders with an undesirable status quo if they do not approve the managerrsquos proposed
action)
References
Associated Press 2007 Administration opposes say on pay bill April 19
Bainbridge S 2008 Remarks on say on pay An unjustified incursion on director authority
UCLA School of Law Research Paper No 08-06
Bebchuk L 2005 The Case for Increasing Shareholder Power Harvard Law Review 118833-
917
Bebchuk L 2007 Written testimony submitted before the Committee on Financial Services
United States House of Representatives Hearing on Empowering Shareholders on Executive
Compensation March 8
Bebchuk LA A Cohen and A Ferrell 2009 What Matters in Corporate Governance Review
of Financial Studies 22783-827
Cai J J Garner and R Walkling 2009 Electing Directors Journal of Finance 642387-2419
Cai J and R Walkling 2011 Shareholdersrsquo Say on Pay Does it Create Value Journal of
Financial and Quantitative Analysis 46299ndash339
Coates IV JC (2009) Testimony before the Subcommittee on Securities Insurance and
Investment of the Committee on Banking Housing and Urban Affairs United States Senate
July 29 2009 httppapersssrncomsol3paperscfmabstract_id=1475355
Correa R and U Lel 2013 Say On Pay Laws Executive Compensation CEO Pay Slice and
Firm Value Around the World Federal Reserve Board Working Paper
Cuntildeat V M Gine and M Guadalupe 2012 The Vote is Cast The Effect of Corporate
Governance on Shareholder Value Journal of Finance 671943-1977
Cuntildeat V M Gine and M Guadalupe 2013 Say Pays Shareholder Voice and Firm
Performance ECGI - Finance Working Paper No 373
Del Guercio D L Seery and T Woidtke 2008 Do Boards Pay Attention When Institutional
Investors lsquoJust Vote Norsquo Journal of Financial Economics 9084-103
Ertimur Y F Ferri and S Stubben 2010 Board of Directorslsquo Responsiveness to Shareholders
Evidence from Shareholder Proposals Journal of Corporate Finance 1653-72
Ertimur Y F Ferri and V Muslu 2011 Shareholder Activism and CEO Pay Review of
Financial Studies 24535ndash592
Ertimur Y F Ferri and D Oesch 2013 Shareholder Votes and Proxy Advisors Evidence from
Say on Pay Journal of Accounting Research 51951-996
Ertimur Y F Ferri and D Oesch forthcoming Does the Director Election System Matter
Evidence from Majority Voting Review of Accounting Studies forthcoming
Ferri F 2012 Low-Cost Shareholder Activism A Review of the Evidence Research
Handbook on the Economics of Corporate Law ed CA Hill and BH McDonnell
Edward Elgar Publishing
Ferri F and D Maber 2013 Say on Pay Votes and CEO Compensation Evidence from the UK
Review of Finance 17527-563
Ferri F and D Oesch 2013 Management Influence on Investors Evidence from Shareholder
Votes on the Frequency of Say on Pay Columbia Business School Research Paper No 13-17
Ferri F and T Sandino 2009 The impact of shareholder activism on financial reporting and
compensation The case of employee stock options expensing The Accounting Review 84433ndash
466
Ferri F and J Weber 2009 AFSCME vs Mozilohellipand Say on Pay for All (A) Harvard
Business School Case 109-009
Gompers P J Ishii and A Metrick 2003 Corporate Governance and Equity Prices Quarterly
Journal of Economics 118107-155
Gordon J 2009 Say on Paylsquo Cautionary Notes on the UK Experience and the Case for
Shareholders Opt-In Harvard Journal on Legislation 46323-368
Iliev P and S Vitanova 2013 The effect of the Say on Pay in the US Pennsylvania State
University Working Paper
Kaplan S 2007 Testimony of Steven N Kaplan on Empowering Shareholders on Executive
Compensation and HR 1257 the Shareholder Vote on Executive Compensation Act before the
Committee on Financial Services United States House of Representatives March 8
Larcker D A McCall and G Ormazabal 2013 The Economic Consequences of Proxy
Advisor Say-on-Pay Voting Policies Working Paper Stanford University
Larcker D G Ormazabal and D Taylor 2011 The market reaction to corporate governance
regulation Journal of Financial Economics 101431ndash448
Levit D and N Malenko 2011 Non-binding voting for shareholder proposals Journal of
Finance 661579ndash1614
Mishel L and N Sabadish 2012 How executive compensation and financial-sector pay have
fueled income inequality Issue Brief 331 Economic Policy Institute
Murphy K 2012 The Politics of Pay A Legislative History of Executive Compensation in
Jennifer Hill and Randall Thomas eds Research Handbook on Executive Pay Edward Elgar
Publishers
New York Times 2007 House votes to give investors say on executive pay April 21
Norris F 2004 Corporate democracy and the power to embarrass New York Times March 4
Ozbas O and J Matsusaka2013 Managerial Accommodation Proxy Access and the Cost of
Shareholder Empowerment University of Southern California CLEO Research Paper Series No
C12-1
Thomas R and J Cotter 2007 Shareholder proposals in the new millennium Shareholder
support board response and market reaction Journal of Corporate Finance 13 368ndash391
Thomas R A Palmiter and J Cotter 2012 Dodd-Franks Say on Pay Will it Lead to a Greater
Role for Shareholders in Corporate Governance Cornell Law Review 971213-1266
Thomas R and C Van der Elst 2013 The International Scope of Say on Pay ECGI Law
Working Paper No227
prospects of the bill in the Senate were uncertain (New York Times 2007) and the Bush White
House openly opposed the Bill (Associated Press 2007) Hence it is not clear the extent to
which this event increased the likelihood of a say on pay legislation Arguably as noted in
Section I the only lsquoeventrsquo that substantially increased the likelihood of say on pay legislation
was the financial crisis and the related perception of compensation abuses (eg outrage over AIG
bonuses)
Ferri and Maber (2013) argue that the announcement of the submission of SOP regulation to the
Parliament in the United Kingdom in June 2002 offers a more powerful setting for an event
study since it was largely unexpected and increased substantially the probability of SOP
adoption (Parliamentrsquos approval was virtually guaranteed) Using this event they document
positive abnormal returns for firms with excess CEO pay combined with poor performance and
for firms with controversial pay practices that weaken the penalties for poor performance (those
practices were often removed in response to SOP votes as discussed earlier in Section IIB)
They interpret their findings as consistent with shareholders perceiving SOP as a value
enhancing monitoring mechanism for firms with low pay-to-poor-performance sensitivity
Finally a recent study by Iliev and Vitanova (2013) examines the market reaction around the
announcement of the SEC final SOP rule that on January 25 2011 exempted small firms (ie
firms with a public float below $75 million) from adopting SOP for two years (press reports at
the time noted that the exemption could become permanent) An appealing feature of this setting
is that the SEC decision was an unexpected reversal from the rule proposed in October 2010 rule
under which no publicly traded firm would be exempted
Iliev and Vitanova (2013) compare the market reaction to this announcement for firms just above
the SEC-imposed threshold and thus subject to the SOP rule and firms just below the threshold
(but otherwise quite similar) and thus exempted from SOP While they find not significant
abnormal returns for either group the abnormal returns for exempted firms are significantly
more negative a finding that the authors interpret as evidence of a positive value effect of
mandatory SOP While the setting is an interesting one some concerns remain as exempted
firms are fairly small and the magnitude of the executive pay problem (if any) is unlikely to be
large enough to explain the differential market reaction Also the same study finds no effect on
CEO pay level and composition (see Section IIA) Hence the reason for the differential stock
price reaction remains unclear It would be helpful to know if the event study result is driven by
(or stronger for) exempted firms with excessive pay or problematic pay practices
B Event studies around Shareholder Proposals to Adopt Say on Pay
Starting in 2006 shareholder activists led by union pension funds submitted shareholder
proposal to adopt SOP at hundreds of US firms in an attempt to induce voluntary or mandatory
widespread adoption of SOP
Cai and Walkling (2011) examine the market reaction to proxy filings and annual meetings of
113 firms targeted by nonbinding shareholder proposals to adopt SOP between 2006 and 2008
On average they find insignificant returns around both the submission of the proposal (proxy
filing date) and the vote on the proposal (annual meeting date) But they also find that when the
proposals are filed by union pension funds the abnormal returns (still insignificant) are more
negative than when filed by other activists and that when the proposal is defeated the abnormal
returns (while insignificant) are more positive than when the proposal is passed (it is not clear
whether this result is driven by union-sponsored proposals) They interpret these findings as
evidence that the market views SOP proposals filed by union pension funds are driven by special
interests rather than value maximization10
However caution is required in interpreting this
evidence Virtually all activists filing SOP proposals were coordinated by a group of investors
(mostly but not only union pension funds) who made available to other activists a template for
SOP proposals and a list of target firms (Ferri and Weber 2009) Hence the distinction between
union and non-union proponents in the context of SOP proposals may be only apparent More
importantly the list of target firms was publicly available months before the proxy filing dates
so it is not clear whether the proxy statements contained any new information Similarly the
voting outcome of the SOP proposals may largely be anticipated (based on the composition of
institutional owners proxy advisorsrsquo recommendations etc) Finally the analyses do not control
for other (potentially new) information contained in proxy statements (eg executive
compensation report other items up for a vote at the annual meeting) or other events occurring at
the annual meeting (eg shareholder votes on other items)
Cuntildeat Gine and Guadalupe (2013) also examine the market reaction to the voting outcome of
nonbinding shareholder proposals to adopt SOP but to alleviate these concerns they employ a
lsquofuzzyrsquo regression discontinuity design (RDD) essentially comparing the stock price reaction to
SOP proposals that pass by a small margin to the reaction to SOP proposals that fail by a small
10
Consistent with this interpretation the authors also find that SOP proposals generally targeted larger firms rather
than firms with excessive pay or lower pay-performance sensitivity However as noted by Ferri and Sandino (2009)
activists typically submit shareholder proposals aimed at promoting policy reforms at a broad sample of large firms
rather than the firms that would most benefit from the proposals because they believe that showing widespread
support for the proposal across all types of firms enhances its credibility with policy makers
margin (Cuntildeat Gine and Guadalupe 2012 Ertimur Ferri and Oesch forthcoming) The underlying idea
is that firms around the threshold are likely to have similar characteristics (indeed they do as the
authors show) but differ in the likelihood of implementation which is typically much higher for
proposals passing the threshold (Ertimur Ferri and Stubben 2010 Thomas and Cotter 2007)
Indeed the authors show that the likelihood of subsequently adopting SOP is 48 higher when
the SOP proposal passes by a small margin relatively to when it fails by a small margin Besides
because in these close-call situations the voting outcome is uncertain its resolution (pass or fail)
is likely to convey new information about the likelihood of SOP adoption making this setting
suitable to an event study11
Using this RDD approach Cuntildeat Gine and Guadalupe (2013) find that on the day of the vote a
SOP proposal that passes by a small margin yields an abnormal return of 24 relative to one
that fails (after controlling for other proposals voted upon at the same meeting) and estimate the
lsquofullrsquo value of SOP at 46 (after taking into account the increase in the probability of SOP
implementation)12
Aside from the issues discussed in Section IA (ie generalizability to other
firms) this estimate seems too large to reflect the present value of future reductions in excess
CEO pay (as noted by the authors their estimate of the value of SOP is equivalent to the
estimated value of removing two anti-takeover provision based on Cuntildeat Gine and Guadalupe
2012) particularly because (i) the authors find no evidence of subsequent changes in levels and
composition of CEO pay for firms implementing SOP (see Section IIA) an (ii) the sample firms
11
Consistent with this observation the authors find no market reaction to the voting outcome of SOP shareholder
proposals away from the threshold
12 Since the outcome of the vote is not binding the 24 market reaction only reflects the expected increase in the
probability of SOP adoption and thus understates the value of the SOP provision
do not appear to be characterized by excess CEO pay However the authors also find significant
improvements in subsequent operating performance and efficiency as a result of passing SOP
proposals and thus conjecture that the SOP vote is viewed as a tool to express a vote of
confidence in management performance more than a way to change compensation practices and
that the large positive market reaction reflects expected performance improvements under this
tighter monitoring regime While this is an interesting idea it remains unclear what additional
ldquoteethrdquo a SOP vote may provide over existing (non-compensation related) monitoring
mechanisms shareholders can express (and do often express) their dissatisfaction with
management performance when voting on director elections (Cai Garner and Walkling 2009)
C Event studies around compensation changes induced by SOP
Two studies examine announcements of changes to compensation plans related to SOP Larcker
McCall and Ormazabal (2013) examine the market reaction to (non-contaminated) compensation
changes disclosed in 8-K filings by Russell 3000 firms in the US during the year before the first
SOP vote They find that changes that appear to be made to avoid a negative proxy advisorrsquos
recommendation and thus a negative SOP vote are associated with a negative abnormal return
of -044 whereas other compensation changes are not associated with a significant stock price
reaction Ertimur Ferri and Oesch (2013) perform a similar analysis but focusing on
compensation changes made after the first SOP vote (and explicitly in response to the vote) by
firms receiving a negative recommendation and a high voting dissent in 2011 They fail to find
significant abnormal returns even for the subset of compensation changes that resulted in a
positive recommendation and low dissent in 2012 (and thus were presumably perceived to be
adequate and material by proxy advisors and voting shareholders)
D Other approaches effect of SOP on Tobinrsquos Q
In their cross-country study Correa and Lel (2013) also examine the effect of SOP laws on
Tobinrsquos Q and report a 36 increase in firm value following the adoption of the SOP laws
(more precisely a 36 higher firm value for post-SOP observations relative to a control sample
that pools together pre-SOP and non-SOP observations) They acknowledge that this increase in
firm value is too large to be justified by the relative decrease in CEO pay that they document and
suggest (but do not test) that it may reflect better alignment of pay and performance13
A key
concern is that higher valuation in countries with SOP laws may reflect other governance
changes introduced at the same time While the study controls for other compensation-related
laws there may be other non-compensation regulations introduced with SOP and with a
potentially larger impact on firm value
IV What have we learned
Adoption of SOP in the US has been long advocated by those who contend that executive pay is
a manifestation of rather than a solution to the agency problem and by those who favor great
shareholder involvement in corporate decisions A growing body of research is starting to
investigate the effect of SOP on executive pay and firm value It is premature to draw definitive
conclusions also because these studies differ in methodologies and settings but three points
seem to emerge First there is robust evidence that boards respond to SOP votes when
13
They also find that the firm value increase is higher when the relative decrease in CEO pay results in a lower pay
differential between CEO and other top managers (CEO pay slice) implying that a source of value creation may be
the reduced pay inequality among the top management team But this seems unlikely to explain the change in
Tobinrsquos Q since the increase occurs for both firms with advisory SOP laws and firms with binding SOP laws and
there is no change in CEO pay slice in firms subject to binding SOP laws
shareholders use it ldquovoicerdquo is heardrdquo Both in the US and UK studies show that not only firms
failing to win the SOP vote but also firms facing substantial dissent (20-30 of the votes
against) make changes to their compensation plans in consultation with institutional investors
and proxy advisors Second in both the US and UK institutional investors appear to use the
power of SOP votes to pressure firms into strengthening the link (or the perceived link) between
pay and performance (particularly on the downside) but not to pressure firms to reduce target
levels of pay suggesting a reluctance of institutional investors to lsquoregulatersquo executive pay Third
and consistent with the above most studies examining the aggregate effect of SOP on CEO pay
find some increase in pay-performance sensitivity but not effect on the level and growth of CEO
pay While some observers may interpret the lack of an effect on CEO pay levels as evidence of
the failure of SOP one should note that SOP per se is a neutral tool Its impact depends on what
investors choose to lsquosay on payrsquo Fourth based on my interpretation of the existing research to
date there is little evidence of an effect of SOP on firm value The studies documenting a
positive price impact appear subject to alternative interpretations or not plausible in their
findings (eg they fail to find an effect on CEO pay or the effect is too small to justify the
documented price impact hence it remains unclear what the source of value creation is) Also
there seems to be no stock price reaction to the SOP-induced compensation changes Finally
given the nature of these SOP-induced changes while some of them may be beneficial it seems
hard to believe that they will have a statistically detectable impact on firm value except perhaps
in a handful of firms where the compensation plan is subject to a complete overhaul
Perhaps one explanation for the weak impact of SOP is that executive pay problems were
overstated or that by the time SOP was introduced (more than a decade after the Enron-type
scandals that led to calls for greater shareholder voice) they had been already addressed via
other mechanisms (hedge fund activism monitoring by institutional investors vote-no
campaigns against compensation committee members SEC-mandated pay disclosures)
Overall though based on the evidence to date it does not appear that SOP had a major impact14
Was the adoption of SOP an lsquooptimalrsquo choice from a social welfare point of view This is a
difficult question to answer given the challenges of identifying and measuring all the potential
costs and benefits associated with SOP (likely to change over time and differ across countries)
Perhaps the most positive view of SOP is that its introduction prevented the adoption of other
more radical and intrusive regulatory measures (eg CEO pay caps) In that sense it may be
viewed as an lsquooptimalrsquo answer (ie the lesser of two evils) to the political pressure to reform
executive pay during the financial crisis
14
This interpretation of the evidence is also consistent with the prediction of analytical studies Ozbas and
Matsusaka (2013) model the benefits and costs of shareholder empowerment distinguishing between the right to
approve and the right to propose They show that permitting shareholders to propose directors or policies can cause
value-maximizing managers to take value-reducing actions to accommodate activist investors with non-value-
maximizing goals As for the right to approve it is weakly beneficial that is approval rights (such as a SOP vote)
are generally beneficial for shareholders in that they limit the managerrsquos ability to pursue private benefits at
shareholder expense but they have minimal impact on managerial actions and firm value (because the manager in
effect can threaten shareholders with an undesirable status quo if they do not approve the managerrsquos proposed
action)
References
Associated Press 2007 Administration opposes say on pay bill April 19
Bainbridge S 2008 Remarks on say on pay An unjustified incursion on director authority
UCLA School of Law Research Paper No 08-06
Bebchuk L 2005 The Case for Increasing Shareholder Power Harvard Law Review 118833-
917
Bebchuk L 2007 Written testimony submitted before the Committee on Financial Services
United States House of Representatives Hearing on Empowering Shareholders on Executive
Compensation March 8
Bebchuk LA A Cohen and A Ferrell 2009 What Matters in Corporate Governance Review
of Financial Studies 22783-827
Cai J J Garner and R Walkling 2009 Electing Directors Journal of Finance 642387-2419
Cai J and R Walkling 2011 Shareholdersrsquo Say on Pay Does it Create Value Journal of
Financial and Quantitative Analysis 46299ndash339
Coates IV JC (2009) Testimony before the Subcommittee on Securities Insurance and
Investment of the Committee on Banking Housing and Urban Affairs United States Senate
July 29 2009 httppapersssrncomsol3paperscfmabstract_id=1475355
Correa R and U Lel 2013 Say On Pay Laws Executive Compensation CEO Pay Slice and
Firm Value Around the World Federal Reserve Board Working Paper
Cuntildeat V M Gine and M Guadalupe 2012 The Vote is Cast The Effect of Corporate
Governance on Shareholder Value Journal of Finance 671943-1977
Cuntildeat V M Gine and M Guadalupe 2013 Say Pays Shareholder Voice and Firm
Performance ECGI - Finance Working Paper No 373
Del Guercio D L Seery and T Woidtke 2008 Do Boards Pay Attention When Institutional
Investors lsquoJust Vote Norsquo Journal of Financial Economics 9084-103
Ertimur Y F Ferri and S Stubben 2010 Board of Directorslsquo Responsiveness to Shareholders
Evidence from Shareholder Proposals Journal of Corporate Finance 1653-72
Ertimur Y F Ferri and V Muslu 2011 Shareholder Activism and CEO Pay Review of
Financial Studies 24535ndash592
Ertimur Y F Ferri and D Oesch 2013 Shareholder Votes and Proxy Advisors Evidence from
Say on Pay Journal of Accounting Research 51951-996
Ertimur Y F Ferri and D Oesch forthcoming Does the Director Election System Matter
Evidence from Majority Voting Review of Accounting Studies forthcoming
Ferri F 2012 Low-Cost Shareholder Activism A Review of the Evidence Research
Handbook on the Economics of Corporate Law ed CA Hill and BH McDonnell
Edward Elgar Publishing
Ferri F and D Maber 2013 Say on Pay Votes and CEO Compensation Evidence from the UK
Review of Finance 17527-563
Ferri F and D Oesch 2013 Management Influence on Investors Evidence from Shareholder
Votes on the Frequency of Say on Pay Columbia Business School Research Paper No 13-17
Ferri F and T Sandino 2009 The impact of shareholder activism on financial reporting and
compensation The case of employee stock options expensing The Accounting Review 84433ndash
466
Ferri F and J Weber 2009 AFSCME vs Mozilohellipand Say on Pay for All (A) Harvard
Business School Case 109-009
Gompers P J Ishii and A Metrick 2003 Corporate Governance and Equity Prices Quarterly
Journal of Economics 118107-155
Gordon J 2009 Say on Paylsquo Cautionary Notes on the UK Experience and the Case for
Shareholders Opt-In Harvard Journal on Legislation 46323-368
Iliev P and S Vitanova 2013 The effect of the Say on Pay in the US Pennsylvania State
University Working Paper
Kaplan S 2007 Testimony of Steven N Kaplan on Empowering Shareholders on Executive
Compensation and HR 1257 the Shareholder Vote on Executive Compensation Act before the
Committee on Financial Services United States House of Representatives March 8
Larcker D A McCall and G Ormazabal 2013 The Economic Consequences of Proxy
Advisor Say-on-Pay Voting Policies Working Paper Stanford University
Larcker D G Ormazabal and D Taylor 2011 The market reaction to corporate governance
regulation Journal of Financial Economics 101431ndash448
Levit D and N Malenko 2011 Non-binding voting for shareholder proposals Journal of
Finance 661579ndash1614
Mishel L and N Sabadish 2012 How executive compensation and financial-sector pay have
fueled income inequality Issue Brief 331 Economic Policy Institute
Murphy K 2012 The Politics of Pay A Legislative History of Executive Compensation in
Jennifer Hill and Randall Thomas eds Research Handbook on Executive Pay Edward Elgar
Publishers
New York Times 2007 House votes to give investors say on executive pay April 21
Norris F 2004 Corporate democracy and the power to embarrass New York Times March 4
Ozbas O and J Matsusaka2013 Managerial Accommodation Proxy Access and the Cost of
Shareholder Empowerment University of Southern California CLEO Research Paper Series No
C12-1
Thomas R and J Cotter 2007 Shareholder proposals in the new millennium Shareholder
support board response and market reaction Journal of Corporate Finance 13 368ndash391
Thomas R A Palmiter and J Cotter 2012 Dodd-Franks Say on Pay Will it Lead to a Greater
Role for Shareholders in Corporate Governance Cornell Law Review 971213-1266
Thomas R and C Van der Elst 2013 The International Scope of Say on Pay ECGI Law
Working Paper No227
Iliev and Vitanova (2013) compare the market reaction to this announcement for firms just above
the SEC-imposed threshold and thus subject to the SOP rule and firms just below the threshold
(but otherwise quite similar) and thus exempted from SOP While they find not significant
abnormal returns for either group the abnormal returns for exempted firms are significantly
more negative a finding that the authors interpret as evidence of a positive value effect of
mandatory SOP While the setting is an interesting one some concerns remain as exempted
firms are fairly small and the magnitude of the executive pay problem (if any) is unlikely to be
large enough to explain the differential market reaction Also the same study finds no effect on
CEO pay level and composition (see Section IIA) Hence the reason for the differential stock
price reaction remains unclear It would be helpful to know if the event study result is driven by
(or stronger for) exempted firms with excessive pay or problematic pay practices
B Event studies around Shareholder Proposals to Adopt Say on Pay
Starting in 2006 shareholder activists led by union pension funds submitted shareholder
proposal to adopt SOP at hundreds of US firms in an attempt to induce voluntary or mandatory
widespread adoption of SOP
Cai and Walkling (2011) examine the market reaction to proxy filings and annual meetings of
113 firms targeted by nonbinding shareholder proposals to adopt SOP between 2006 and 2008
On average they find insignificant returns around both the submission of the proposal (proxy
filing date) and the vote on the proposal (annual meeting date) But they also find that when the
proposals are filed by union pension funds the abnormal returns (still insignificant) are more
negative than when filed by other activists and that when the proposal is defeated the abnormal
returns (while insignificant) are more positive than when the proposal is passed (it is not clear
whether this result is driven by union-sponsored proposals) They interpret these findings as
evidence that the market views SOP proposals filed by union pension funds are driven by special
interests rather than value maximization10
However caution is required in interpreting this
evidence Virtually all activists filing SOP proposals were coordinated by a group of investors
(mostly but not only union pension funds) who made available to other activists a template for
SOP proposals and a list of target firms (Ferri and Weber 2009) Hence the distinction between
union and non-union proponents in the context of SOP proposals may be only apparent More
importantly the list of target firms was publicly available months before the proxy filing dates
so it is not clear whether the proxy statements contained any new information Similarly the
voting outcome of the SOP proposals may largely be anticipated (based on the composition of
institutional owners proxy advisorsrsquo recommendations etc) Finally the analyses do not control
for other (potentially new) information contained in proxy statements (eg executive
compensation report other items up for a vote at the annual meeting) or other events occurring at
the annual meeting (eg shareholder votes on other items)
Cuntildeat Gine and Guadalupe (2013) also examine the market reaction to the voting outcome of
nonbinding shareholder proposals to adopt SOP but to alleviate these concerns they employ a
lsquofuzzyrsquo regression discontinuity design (RDD) essentially comparing the stock price reaction to
SOP proposals that pass by a small margin to the reaction to SOP proposals that fail by a small
10
Consistent with this interpretation the authors also find that SOP proposals generally targeted larger firms rather
than firms with excessive pay or lower pay-performance sensitivity However as noted by Ferri and Sandino (2009)
activists typically submit shareholder proposals aimed at promoting policy reforms at a broad sample of large firms
rather than the firms that would most benefit from the proposals because they believe that showing widespread
support for the proposal across all types of firms enhances its credibility with policy makers
margin (Cuntildeat Gine and Guadalupe 2012 Ertimur Ferri and Oesch forthcoming) The underlying idea
is that firms around the threshold are likely to have similar characteristics (indeed they do as the
authors show) but differ in the likelihood of implementation which is typically much higher for
proposals passing the threshold (Ertimur Ferri and Stubben 2010 Thomas and Cotter 2007)
Indeed the authors show that the likelihood of subsequently adopting SOP is 48 higher when
the SOP proposal passes by a small margin relatively to when it fails by a small margin Besides
because in these close-call situations the voting outcome is uncertain its resolution (pass or fail)
is likely to convey new information about the likelihood of SOP adoption making this setting
suitable to an event study11
Using this RDD approach Cuntildeat Gine and Guadalupe (2013) find that on the day of the vote a
SOP proposal that passes by a small margin yields an abnormal return of 24 relative to one
that fails (after controlling for other proposals voted upon at the same meeting) and estimate the
lsquofullrsquo value of SOP at 46 (after taking into account the increase in the probability of SOP
implementation)12
Aside from the issues discussed in Section IA (ie generalizability to other
firms) this estimate seems too large to reflect the present value of future reductions in excess
CEO pay (as noted by the authors their estimate of the value of SOP is equivalent to the
estimated value of removing two anti-takeover provision based on Cuntildeat Gine and Guadalupe
2012) particularly because (i) the authors find no evidence of subsequent changes in levels and
composition of CEO pay for firms implementing SOP (see Section IIA) an (ii) the sample firms
11
Consistent with this observation the authors find no market reaction to the voting outcome of SOP shareholder
proposals away from the threshold
12 Since the outcome of the vote is not binding the 24 market reaction only reflects the expected increase in the
probability of SOP adoption and thus understates the value of the SOP provision
do not appear to be characterized by excess CEO pay However the authors also find significant
improvements in subsequent operating performance and efficiency as a result of passing SOP
proposals and thus conjecture that the SOP vote is viewed as a tool to express a vote of
confidence in management performance more than a way to change compensation practices and
that the large positive market reaction reflects expected performance improvements under this
tighter monitoring regime While this is an interesting idea it remains unclear what additional
ldquoteethrdquo a SOP vote may provide over existing (non-compensation related) monitoring
mechanisms shareholders can express (and do often express) their dissatisfaction with
management performance when voting on director elections (Cai Garner and Walkling 2009)
C Event studies around compensation changes induced by SOP
Two studies examine announcements of changes to compensation plans related to SOP Larcker
McCall and Ormazabal (2013) examine the market reaction to (non-contaminated) compensation
changes disclosed in 8-K filings by Russell 3000 firms in the US during the year before the first
SOP vote They find that changes that appear to be made to avoid a negative proxy advisorrsquos
recommendation and thus a negative SOP vote are associated with a negative abnormal return
of -044 whereas other compensation changes are not associated with a significant stock price
reaction Ertimur Ferri and Oesch (2013) perform a similar analysis but focusing on
compensation changes made after the first SOP vote (and explicitly in response to the vote) by
firms receiving a negative recommendation and a high voting dissent in 2011 They fail to find
significant abnormal returns even for the subset of compensation changes that resulted in a
positive recommendation and low dissent in 2012 (and thus were presumably perceived to be
adequate and material by proxy advisors and voting shareholders)
D Other approaches effect of SOP on Tobinrsquos Q
In their cross-country study Correa and Lel (2013) also examine the effect of SOP laws on
Tobinrsquos Q and report a 36 increase in firm value following the adoption of the SOP laws
(more precisely a 36 higher firm value for post-SOP observations relative to a control sample
that pools together pre-SOP and non-SOP observations) They acknowledge that this increase in
firm value is too large to be justified by the relative decrease in CEO pay that they document and
suggest (but do not test) that it may reflect better alignment of pay and performance13
A key
concern is that higher valuation in countries with SOP laws may reflect other governance
changes introduced at the same time While the study controls for other compensation-related
laws there may be other non-compensation regulations introduced with SOP and with a
potentially larger impact on firm value
IV What have we learned
Adoption of SOP in the US has been long advocated by those who contend that executive pay is
a manifestation of rather than a solution to the agency problem and by those who favor great
shareholder involvement in corporate decisions A growing body of research is starting to
investigate the effect of SOP on executive pay and firm value It is premature to draw definitive
conclusions also because these studies differ in methodologies and settings but three points
seem to emerge First there is robust evidence that boards respond to SOP votes when
13
They also find that the firm value increase is higher when the relative decrease in CEO pay results in a lower pay
differential between CEO and other top managers (CEO pay slice) implying that a source of value creation may be
the reduced pay inequality among the top management team But this seems unlikely to explain the change in
Tobinrsquos Q since the increase occurs for both firms with advisory SOP laws and firms with binding SOP laws and
there is no change in CEO pay slice in firms subject to binding SOP laws
shareholders use it ldquovoicerdquo is heardrdquo Both in the US and UK studies show that not only firms
failing to win the SOP vote but also firms facing substantial dissent (20-30 of the votes
against) make changes to their compensation plans in consultation with institutional investors
and proxy advisors Second in both the US and UK institutional investors appear to use the
power of SOP votes to pressure firms into strengthening the link (or the perceived link) between
pay and performance (particularly on the downside) but not to pressure firms to reduce target
levels of pay suggesting a reluctance of institutional investors to lsquoregulatersquo executive pay Third
and consistent with the above most studies examining the aggregate effect of SOP on CEO pay
find some increase in pay-performance sensitivity but not effect on the level and growth of CEO
pay While some observers may interpret the lack of an effect on CEO pay levels as evidence of
the failure of SOP one should note that SOP per se is a neutral tool Its impact depends on what
investors choose to lsquosay on payrsquo Fourth based on my interpretation of the existing research to
date there is little evidence of an effect of SOP on firm value The studies documenting a
positive price impact appear subject to alternative interpretations or not plausible in their
findings (eg they fail to find an effect on CEO pay or the effect is too small to justify the
documented price impact hence it remains unclear what the source of value creation is) Also
there seems to be no stock price reaction to the SOP-induced compensation changes Finally
given the nature of these SOP-induced changes while some of them may be beneficial it seems
hard to believe that they will have a statistically detectable impact on firm value except perhaps
in a handful of firms where the compensation plan is subject to a complete overhaul
Perhaps one explanation for the weak impact of SOP is that executive pay problems were
overstated or that by the time SOP was introduced (more than a decade after the Enron-type
scandals that led to calls for greater shareholder voice) they had been already addressed via
other mechanisms (hedge fund activism monitoring by institutional investors vote-no
campaigns against compensation committee members SEC-mandated pay disclosures)
Overall though based on the evidence to date it does not appear that SOP had a major impact14
Was the adoption of SOP an lsquooptimalrsquo choice from a social welfare point of view This is a
difficult question to answer given the challenges of identifying and measuring all the potential
costs and benefits associated with SOP (likely to change over time and differ across countries)
Perhaps the most positive view of SOP is that its introduction prevented the adoption of other
more radical and intrusive regulatory measures (eg CEO pay caps) In that sense it may be
viewed as an lsquooptimalrsquo answer (ie the lesser of two evils) to the political pressure to reform
executive pay during the financial crisis
14
This interpretation of the evidence is also consistent with the prediction of analytical studies Ozbas and
Matsusaka (2013) model the benefits and costs of shareholder empowerment distinguishing between the right to
approve and the right to propose They show that permitting shareholders to propose directors or policies can cause
value-maximizing managers to take value-reducing actions to accommodate activist investors with non-value-
maximizing goals As for the right to approve it is weakly beneficial that is approval rights (such as a SOP vote)
are generally beneficial for shareholders in that they limit the managerrsquos ability to pursue private benefits at
shareholder expense but they have minimal impact on managerial actions and firm value (because the manager in
effect can threaten shareholders with an undesirable status quo if they do not approve the managerrsquos proposed
action)
References
Associated Press 2007 Administration opposes say on pay bill April 19
Bainbridge S 2008 Remarks on say on pay An unjustified incursion on director authority
UCLA School of Law Research Paper No 08-06
Bebchuk L 2005 The Case for Increasing Shareholder Power Harvard Law Review 118833-
917
Bebchuk L 2007 Written testimony submitted before the Committee on Financial Services
United States House of Representatives Hearing on Empowering Shareholders on Executive
Compensation March 8
Bebchuk LA A Cohen and A Ferrell 2009 What Matters in Corporate Governance Review
of Financial Studies 22783-827
Cai J J Garner and R Walkling 2009 Electing Directors Journal of Finance 642387-2419
Cai J and R Walkling 2011 Shareholdersrsquo Say on Pay Does it Create Value Journal of
Financial and Quantitative Analysis 46299ndash339
Coates IV JC (2009) Testimony before the Subcommittee on Securities Insurance and
Investment of the Committee on Banking Housing and Urban Affairs United States Senate
July 29 2009 httppapersssrncomsol3paperscfmabstract_id=1475355
Correa R and U Lel 2013 Say On Pay Laws Executive Compensation CEO Pay Slice and
Firm Value Around the World Federal Reserve Board Working Paper
Cuntildeat V M Gine and M Guadalupe 2012 The Vote is Cast The Effect of Corporate
Governance on Shareholder Value Journal of Finance 671943-1977
Cuntildeat V M Gine and M Guadalupe 2013 Say Pays Shareholder Voice and Firm
Performance ECGI - Finance Working Paper No 373
Del Guercio D L Seery and T Woidtke 2008 Do Boards Pay Attention When Institutional
Investors lsquoJust Vote Norsquo Journal of Financial Economics 9084-103
Ertimur Y F Ferri and S Stubben 2010 Board of Directorslsquo Responsiveness to Shareholders
Evidence from Shareholder Proposals Journal of Corporate Finance 1653-72
Ertimur Y F Ferri and V Muslu 2011 Shareholder Activism and CEO Pay Review of
Financial Studies 24535ndash592
Ertimur Y F Ferri and D Oesch 2013 Shareholder Votes and Proxy Advisors Evidence from
Say on Pay Journal of Accounting Research 51951-996
Ertimur Y F Ferri and D Oesch forthcoming Does the Director Election System Matter
Evidence from Majority Voting Review of Accounting Studies forthcoming
Ferri F 2012 Low-Cost Shareholder Activism A Review of the Evidence Research
Handbook on the Economics of Corporate Law ed CA Hill and BH McDonnell
Edward Elgar Publishing
Ferri F and D Maber 2013 Say on Pay Votes and CEO Compensation Evidence from the UK
Review of Finance 17527-563
Ferri F and D Oesch 2013 Management Influence on Investors Evidence from Shareholder
Votes on the Frequency of Say on Pay Columbia Business School Research Paper No 13-17
Ferri F and T Sandino 2009 The impact of shareholder activism on financial reporting and
compensation The case of employee stock options expensing The Accounting Review 84433ndash
466
Ferri F and J Weber 2009 AFSCME vs Mozilohellipand Say on Pay for All (A) Harvard
Business School Case 109-009
Gompers P J Ishii and A Metrick 2003 Corporate Governance and Equity Prices Quarterly
Journal of Economics 118107-155
Gordon J 2009 Say on Paylsquo Cautionary Notes on the UK Experience and the Case for
Shareholders Opt-In Harvard Journal on Legislation 46323-368
Iliev P and S Vitanova 2013 The effect of the Say on Pay in the US Pennsylvania State
University Working Paper
Kaplan S 2007 Testimony of Steven N Kaplan on Empowering Shareholders on Executive
Compensation and HR 1257 the Shareholder Vote on Executive Compensation Act before the
Committee on Financial Services United States House of Representatives March 8
Larcker D A McCall and G Ormazabal 2013 The Economic Consequences of Proxy
Advisor Say-on-Pay Voting Policies Working Paper Stanford University
Larcker D G Ormazabal and D Taylor 2011 The market reaction to corporate governance
regulation Journal of Financial Economics 101431ndash448
Levit D and N Malenko 2011 Non-binding voting for shareholder proposals Journal of
Finance 661579ndash1614
Mishel L and N Sabadish 2012 How executive compensation and financial-sector pay have
fueled income inequality Issue Brief 331 Economic Policy Institute
Murphy K 2012 The Politics of Pay A Legislative History of Executive Compensation in
Jennifer Hill and Randall Thomas eds Research Handbook on Executive Pay Edward Elgar
Publishers
New York Times 2007 House votes to give investors say on executive pay April 21
Norris F 2004 Corporate democracy and the power to embarrass New York Times March 4
Ozbas O and J Matsusaka2013 Managerial Accommodation Proxy Access and the Cost of
Shareholder Empowerment University of Southern California CLEO Research Paper Series No
C12-1
Thomas R and J Cotter 2007 Shareholder proposals in the new millennium Shareholder
support board response and market reaction Journal of Corporate Finance 13 368ndash391
Thomas R A Palmiter and J Cotter 2012 Dodd-Franks Say on Pay Will it Lead to a Greater
Role for Shareholders in Corporate Governance Cornell Law Review 971213-1266
Thomas R and C Van der Elst 2013 The International Scope of Say on Pay ECGI Law
Working Paper No227
whether this result is driven by union-sponsored proposals) They interpret these findings as
evidence that the market views SOP proposals filed by union pension funds are driven by special
interests rather than value maximization10
However caution is required in interpreting this
evidence Virtually all activists filing SOP proposals were coordinated by a group of investors
(mostly but not only union pension funds) who made available to other activists a template for
SOP proposals and a list of target firms (Ferri and Weber 2009) Hence the distinction between
union and non-union proponents in the context of SOP proposals may be only apparent More
importantly the list of target firms was publicly available months before the proxy filing dates
so it is not clear whether the proxy statements contained any new information Similarly the
voting outcome of the SOP proposals may largely be anticipated (based on the composition of
institutional owners proxy advisorsrsquo recommendations etc) Finally the analyses do not control
for other (potentially new) information contained in proxy statements (eg executive
compensation report other items up for a vote at the annual meeting) or other events occurring at
the annual meeting (eg shareholder votes on other items)
Cuntildeat Gine and Guadalupe (2013) also examine the market reaction to the voting outcome of
nonbinding shareholder proposals to adopt SOP but to alleviate these concerns they employ a
lsquofuzzyrsquo regression discontinuity design (RDD) essentially comparing the stock price reaction to
SOP proposals that pass by a small margin to the reaction to SOP proposals that fail by a small
10
Consistent with this interpretation the authors also find that SOP proposals generally targeted larger firms rather
than firms with excessive pay or lower pay-performance sensitivity However as noted by Ferri and Sandino (2009)
activists typically submit shareholder proposals aimed at promoting policy reforms at a broad sample of large firms
rather than the firms that would most benefit from the proposals because they believe that showing widespread
support for the proposal across all types of firms enhances its credibility with policy makers
margin (Cuntildeat Gine and Guadalupe 2012 Ertimur Ferri and Oesch forthcoming) The underlying idea
is that firms around the threshold are likely to have similar characteristics (indeed they do as the
authors show) but differ in the likelihood of implementation which is typically much higher for
proposals passing the threshold (Ertimur Ferri and Stubben 2010 Thomas and Cotter 2007)
Indeed the authors show that the likelihood of subsequently adopting SOP is 48 higher when
the SOP proposal passes by a small margin relatively to when it fails by a small margin Besides
because in these close-call situations the voting outcome is uncertain its resolution (pass or fail)
is likely to convey new information about the likelihood of SOP adoption making this setting
suitable to an event study11
Using this RDD approach Cuntildeat Gine and Guadalupe (2013) find that on the day of the vote a
SOP proposal that passes by a small margin yields an abnormal return of 24 relative to one
that fails (after controlling for other proposals voted upon at the same meeting) and estimate the
lsquofullrsquo value of SOP at 46 (after taking into account the increase in the probability of SOP
implementation)12
Aside from the issues discussed in Section IA (ie generalizability to other
firms) this estimate seems too large to reflect the present value of future reductions in excess
CEO pay (as noted by the authors their estimate of the value of SOP is equivalent to the
estimated value of removing two anti-takeover provision based on Cuntildeat Gine and Guadalupe
2012) particularly because (i) the authors find no evidence of subsequent changes in levels and
composition of CEO pay for firms implementing SOP (see Section IIA) an (ii) the sample firms
11
Consistent with this observation the authors find no market reaction to the voting outcome of SOP shareholder
proposals away from the threshold
12 Since the outcome of the vote is not binding the 24 market reaction only reflects the expected increase in the
probability of SOP adoption and thus understates the value of the SOP provision
do not appear to be characterized by excess CEO pay However the authors also find significant
improvements in subsequent operating performance and efficiency as a result of passing SOP
proposals and thus conjecture that the SOP vote is viewed as a tool to express a vote of
confidence in management performance more than a way to change compensation practices and
that the large positive market reaction reflects expected performance improvements under this
tighter monitoring regime While this is an interesting idea it remains unclear what additional
ldquoteethrdquo a SOP vote may provide over existing (non-compensation related) monitoring
mechanisms shareholders can express (and do often express) their dissatisfaction with
management performance when voting on director elections (Cai Garner and Walkling 2009)
C Event studies around compensation changes induced by SOP
Two studies examine announcements of changes to compensation plans related to SOP Larcker
McCall and Ormazabal (2013) examine the market reaction to (non-contaminated) compensation
changes disclosed in 8-K filings by Russell 3000 firms in the US during the year before the first
SOP vote They find that changes that appear to be made to avoid a negative proxy advisorrsquos
recommendation and thus a negative SOP vote are associated with a negative abnormal return
of -044 whereas other compensation changes are not associated with a significant stock price
reaction Ertimur Ferri and Oesch (2013) perform a similar analysis but focusing on
compensation changes made after the first SOP vote (and explicitly in response to the vote) by
firms receiving a negative recommendation and a high voting dissent in 2011 They fail to find
significant abnormal returns even for the subset of compensation changes that resulted in a
positive recommendation and low dissent in 2012 (and thus were presumably perceived to be
adequate and material by proxy advisors and voting shareholders)
D Other approaches effect of SOP on Tobinrsquos Q
In their cross-country study Correa and Lel (2013) also examine the effect of SOP laws on
Tobinrsquos Q and report a 36 increase in firm value following the adoption of the SOP laws
(more precisely a 36 higher firm value for post-SOP observations relative to a control sample
that pools together pre-SOP and non-SOP observations) They acknowledge that this increase in
firm value is too large to be justified by the relative decrease in CEO pay that they document and
suggest (but do not test) that it may reflect better alignment of pay and performance13
A key
concern is that higher valuation in countries with SOP laws may reflect other governance
changes introduced at the same time While the study controls for other compensation-related
laws there may be other non-compensation regulations introduced with SOP and with a
potentially larger impact on firm value
IV What have we learned
Adoption of SOP in the US has been long advocated by those who contend that executive pay is
a manifestation of rather than a solution to the agency problem and by those who favor great
shareholder involvement in corporate decisions A growing body of research is starting to
investigate the effect of SOP on executive pay and firm value It is premature to draw definitive
conclusions also because these studies differ in methodologies and settings but three points
seem to emerge First there is robust evidence that boards respond to SOP votes when
13
They also find that the firm value increase is higher when the relative decrease in CEO pay results in a lower pay
differential between CEO and other top managers (CEO pay slice) implying that a source of value creation may be
the reduced pay inequality among the top management team But this seems unlikely to explain the change in
Tobinrsquos Q since the increase occurs for both firms with advisory SOP laws and firms with binding SOP laws and
there is no change in CEO pay slice in firms subject to binding SOP laws
shareholders use it ldquovoicerdquo is heardrdquo Both in the US and UK studies show that not only firms
failing to win the SOP vote but also firms facing substantial dissent (20-30 of the votes
against) make changes to their compensation plans in consultation with institutional investors
and proxy advisors Second in both the US and UK institutional investors appear to use the
power of SOP votes to pressure firms into strengthening the link (or the perceived link) between
pay and performance (particularly on the downside) but not to pressure firms to reduce target
levels of pay suggesting a reluctance of institutional investors to lsquoregulatersquo executive pay Third
and consistent with the above most studies examining the aggregate effect of SOP on CEO pay
find some increase in pay-performance sensitivity but not effect on the level and growth of CEO
pay While some observers may interpret the lack of an effect on CEO pay levels as evidence of
the failure of SOP one should note that SOP per se is a neutral tool Its impact depends on what
investors choose to lsquosay on payrsquo Fourth based on my interpretation of the existing research to
date there is little evidence of an effect of SOP on firm value The studies documenting a
positive price impact appear subject to alternative interpretations or not plausible in their
findings (eg they fail to find an effect on CEO pay or the effect is too small to justify the
documented price impact hence it remains unclear what the source of value creation is) Also
there seems to be no stock price reaction to the SOP-induced compensation changes Finally
given the nature of these SOP-induced changes while some of them may be beneficial it seems
hard to believe that they will have a statistically detectable impact on firm value except perhaps
in a handful of firms where the compensation plan is subject to a complete overhaul
Perhaps one explanation for the weak impact of SOP is that executive pay problems were
overstated or that by the time SOP was introduced (more than a decade after the Enron-type
scandals that led to calls for greater shareholder voice) they had been already addressed via
other mechanisms (hedge fund activism monitoring by institutional investors vote-no
campaigns against compensation committee members SEC-mandated pay disclosures)
Overall though based on the evidence to date it does not appear that SOP had a major impact14
Was the adoption of SOP an lsquooptimalrsquo choice from a social welfare point of view This is a
difficult question to answer given the challenges of identifying and measuring all the potential
costs and benefits associated with SOP (likely to change over time and differ across countries)
Perhaps the most positive view of SOP is that its introduction prevented the adoption of other
more radical and intrusive regulatory measures (eg CEO pay caps) In that sense it may be
viewed as an lsquooptimalrsquo answer (ie the lesser of two evils) to the political pressure to reform
executive pay during the financial crisis
14
This interpretation of the evidence is also consistent with the prediction of analytical studies Ozbas and
Matsusaka (2013) model the benefits and costs of shareholder empowerment distinguishing between the right to
approve and the right to propose They show that permitting shareholders to propose directors or policies can cause
value-maximizing managers to take value-reducing actions to accommodate activist investors with non-value-
maximizing goals As for the right to approve it is weakly beneficial that is approval rights (such as a SOP vote)
are generally beneficial for shareholders in that they limit the managerrsquos ability to pursue private benefits at
shareholder expense but they have minimal impact on managerial actions and firm value (because the manager in
effect can threaten shareholders with an undesirable status quo if they do not approve the managerrsquos proposed
action)
References
Associated Press 2007 Administration opposes say on pay bill April 19
Bainbridge S 2008 Remarks on say on pay An unjustified incursion on director authority
UCLA School of Law Research Paper No 08-06
Bebchuk L 2005 The Case for Increasing Shareholder Power Harvard Law Review 118833-
917
Bebchuk L 2007 Written testimony submitted before the Committee on Financial Services
United States House of Representatives Hearing on Empowering Shareholders on Executive
Compensation March 8
Bebchuk LA A Cohen and A Ferrell 2009 What Matters in Corporate Governance Review
of Financial Studies 22783-827
Cai J J Garner and R Walkling 2009 Electing Directors Journal of Finance 642387-2419
Cai J and R Walkling 2011 Shareholdersrsquo Say on Pay Does it Create Value Journal of
Financial and Quantitative Analysis 46299ndash339
Coates IV JC (2009) Testimony before the Subcommittee on Securities Insurance and
Investment of the Committee on Banking Housing and Urban Affairs United States Senate
July 29 2009 httppapersssrncomsol3paperscfmabstract_id=1475355
Correa R and U Lel 2013 Say On Pay Laws Executive Compensation CEO Pay Slice and
Firm Value Around the World Federal Reserve Board Working Paper
Cuntildeat V M Gine and M Guadalupe 2012 The Vote is Cast The Effect of Corporate
Governance on Shareholder Value Journal of Finance 671943-1977
Cuntildeat V M Gine and M Guadalupe 2013 Say Pays Shareholder Voice and Firm
Performance ECGI - Finance Working Paper No 373
Del Guercio D L Seery and T Woidtke 2008 Do Boards Pay Attention When Institutional
Investors lsquoJust Vote Norsquo Journal of Financial Economics 9084-103
Ertimur Y F Ferri and S Stubben 2010 Board of Directorslsquo Responsiveness to Shareholders
Evidence from Shareholder Proposals Journal of Corporate Finance 1653-72
Ertimur Y F Ferri and V Muslu 2011 Shareholder Activism and CEO Pay Review of
Financial Studies 24535ndash592
Ertimur Y F Ferri and D Oesch 2013 Shareholder Votes and Proxy Advisors Evidence from
Say on Pay Journal of Accounting Research 51951-996
Ertimur Y F Ferri and D Oesch forthcoming Does the Director Election System Matter
Evidence from Majority Voting Review of Accounting Studies forthcoming
Ferri F 2012 Low-Cost Shareholder Activism A Review of the Evidence Research
Handbook on the Economics of Corporate Law ed CA Hill and BH McDonnell
Edward Elgar Publishing
Ferri F and D Maber 2013 Say on Pay Votes and CEO Compensation Evidence from the UK
Review of Finance 17527-563
Ferri F and D Oesch 2013 Management Influence on Investors Evidence from Shareholder
Votes on the Frequency of Say on Pay Columbia Business School Research Paper No 13-17
Ferri F and T Sandino 2009 The impact of shareholder activism on financial reporting and
compensation The case of employee stock options expensing The Accounting Review 84433ndash
466
Ferri F and J Weber 2009 AFSCME vs Mozilohellipand Say on Pay for All (A) Harvard
Business School Case 109-009
Gompers P J Ishii and A Metrick 2003 Corporate Governance and Equity Prices Quarterly
Journal of Economics 118107-155
Gordon J 2009 Say on Paylsquo Cautionary Notes on the UK Experience and the Case for
Shareholders Opt-In Harvard Journal on Legislation 46323-368
Iliev P and S Vitanova 2013 The effect of the Say on Pay in the US Pennsylvania State
University Working Paper
Kaplan S 2007 Testimony of Steven N Kaplan on Empowering Shareholders on Executive
Compensation and HR 1257 the Shareholder Vote on Executive Compensation Act before the
Committee on Financial Services United States House of Representatives March 8
Larcker D A McCall and G Ormazabal 2013 The Economic Consequences of Proxy
Advisor Say-on-Pay Voting Policies Working Paper Stanford University
Larcker D G Ormazabal and D Taylor 2011 The market reaction to corporate governance
regulation Journal of Financial Economics 101431ndash448
Levit D and N Malenko 2011 Non-binding voting for shareholder proposals Journal of
Finance 661579ndash1614
Mishel L and N Sabadish 2012 How executive compensation and financial-sector pay have
fueled income inequality Issue Brief 331 Economic Policy Institute
Murphy K 2012 The Politics of Pay A Legislative History of Executive Compensation in
Jennifer Hill and Randall Thomas eds Research Handbook on Executive Pay Edward Elgar
Publishers
New York Times 2007 House votes to give investors say on executive pay April 21
Norris F 2004 Corporate democracy and the power to embarrass New York Times March 4
Ozbas O and J Matsusaka2013 Managerial Accommodation Proxy Access and the Cost of
Shareholder Empowerment University of Southern California CLEO Research Paper Series No
C12-1
Thomas R and J Cotter 2007 Shareholder proposals in the new millennium Shareholder
support board response and market reaction Journal of Corporate Finance 13 368ndash391
Thomas R A Palmiter and J Cotter 2012 Dodd-Franks Say on Pay Will it Lead to a Greater
Role for Shareholders in Corporate Governance Cornell Law Review 971213-1266
Thomas R and C Van der Elst 2013 The International Scope of Say on Pay ECGI Law
Working Paper No227
margin (Cuntildeat Gine and Guadalupe 2012 Ertimur Ferri and Oesch forthcoming) The underlying idea
is that firms around the threshold are likely to have similar characteristics (indeed they do as the
authors show) but differ in the likelihood of implementation which is typically much higher for
proposals passing the threshold (Ertimur Ferri and Stubben 2010 Thomas and Cotter 2007)
Indeed the authors show that the likelihood of subsequently adopting SOP is 48 higher when
the SOP proposal passes by a small margin relatively to when it fails by a small margin Besides
because in these close-call situations the voting outcome is uncertain its resolution (pass or fail)
is likely to convey new information about the likelihood of SOP adoption making this setting
suitable to an event study11
Using this RDD approach Cuntildeat Gine and Guadalupe (2013) find that on the day of the vote a
SOP proposal that passes by a small margin yields an abnormal return of 24 relative to one
that fails (after controlling for other proposals voted upon at the same meeting) and estimate the
lsquofullrsquo value of SOP at 46 (after taking into account the increase in the probability of SOP
implementation)12
Aside from the issues discussed in Section IA (ie generalizability to other
firms) this estimate seems too large to reflect the present value of future reductions in excess
CEO pay (as noted by the authors their estimate of the value of SOP is equivalent to the
estimated value of removing two anti-takeover provision based on Cuntildeat Gine and Guadalupe
2012) particularly because (i) the authors find no evidence of subsequent changes in levels and
composition of CEO pay for firms implementing SOP (see Section IIA) an (ii) the sample firms
11
Consistent with this observation the authors find no market reaction to the voting outcome of SOP shareholder
proposals away from the threshold
12 Since the outcome of the vote is not binding the 24 market reaction only reflects the expected increase in the
probability of SOP adoption and thus understates the value of the SOP provision
do not appear to be characterized by excess CEO pay However the authors also find significant
improvements in subsequent operating performance and efficiency as a result of passing SOP
proposals and thus conjecture that the SOP vote is viewed as a tool to express a vote of
confidence in management performance more than a way to change compensation practices and
that the large positive market reaction reflects expected performance improvements under this
tighter monitoring regime While this is an interesting idea it remains unclear what additional
ldquoteethrdquo a SOP vote may provide over existing (non-compensation related) monitoring
mechanisms shareholders can express (and do often express) their dissatisfaction with
management performance when voting on director elections (Cai Garner and Walkling 2009)
C Event studies around compensation changes induced by SOP
Two studies examine announcements of changes to compensation plans related to SOP Larcker
McCall and Ormazabal (2013) examine the market reaction to (non-contaminated) compensation
changes disclosed in 8-K filings by Russell 3000 firms in the US during the year before the first
SOP vote They find that changes that appear to be made to avoid a negative proxy advisorrsquos
recommendation and thus a negative SOP vote are associated with a negative abnormal return
of -044 whereas other compensation changes are not associated with a significant stock price
reaction Ertimur Ferri and Oesch (2013) perform a similar analysis but focusing on
compensation changes made after the first SOP vote (and explicitly in response to the vote) by
firms receiving a negative recommendation and a high voting dissent in 2011 They fail to find
significant abnormal returns even for the subset of compensation changes that resulted in a
positive recommendation and low dissent in 2012 (and thus were presumably perceived to be
adequate and material by proxy advisors and voting shareholders)
D Other approaches effect of SOP on Tobinrsquos Q
In their cross-country study Correa and Lel (2013) also examine the effect of SOP laws on
Tobinrsquos Q and report a 36 increase in firm value following the adoption of the SOP laws
(more precisely a 36 higher firm value for post-SOP observations relative to a control sample
that pools together pre-SOP and non-SOP observations) They acknowledge that this increase in
firm value is too large to be justified by the relative decrease in CEO pay that they document and
suggest (but do not test) that it may reflect better alignment of pay and performance13
A key
concern is that higher valuation in countries with SOP laws may reflect other governance
changes introduced at the same time While the study controls for other compensation-related
laws there may be other non-compensation regulations introduced with SOP and with a
potentially larger impact on firm value
IV What have we learned
Adoption of SOP in the US has been long advocated by those who contend that executive pay is
a manifestation of rather than a solution to the agency problem and by those who favor great
shareholder involvement in corporate decisions A growing body of research is starting to
investigate the effect of SOP on executive pay and firm value It is premature to draw definitive
conclusions also because these studies differ in methodologies and settings but three points
seem to emerge First there is robust evidence that boards respond to SOP votes when
13
They also find that the firm value increase is higher when the relative decrease in CEO pay results in a lower pay
differential between CEO and other top managers (CEO pay slice) implying that a source of value creation may be
the reduced pay inequality among the top management team But this seems unlikely to explain the change in
Tobinrsquos Q since the increase occurs for both firms with advisory SOP laws and firms with binding SOP laws and
there is no change in CEO pay slice in firms subject to binding SOP laws
shareholders use it ldquovoicerdquo is heardrdquo Both in the US and UK studies show that not only firms
failing to win the SOP vote but also firms facing substantial dissent (20-30 of the votes
against) make changes to their compensation plans in consultation with institutional investors
and proxy advisors Second in both the US and UK institutional investors appear to use the
power of SOP votes to pressure firms into strengthening the link (or the perceived link) between
pay and performance (particularly on the downside) but not to pressure firms to reduce target
levels of pay suggesting a reluctance of institutional investors to lsquoregulatersquo executive pay Third
and consistent with the above most studies examining the aggregate effect of SOP on CEO pay
find some increase in pay-performance sensitivity but not effect on the level and growth of CEO
pay While some observers may interpret the lack of an effect on CEO pay levels as evidence of
the failure of SOP one should note that SOP per se is a neutral tool Its impact depends on what
investors choose to lsquosay on payrsquo Fourth based on my interpretation of the existing research to
date there is little evidence of an effect of SOP on firm value The studies documenting a
positive price impact appear subject to alternative interpretations or not plausible in their
findings (eg they fail to find an effect on CEO pay or the effect is too small to justify the
documented price impact hence it remains unclear what the source of value creation is) Also
there seems to be no stock price reaction to the SOP-induced compensation changes Finally
given the nature of these SOP-induced changes while some of them may be beneficial it seems
hard to believe that they will have a statistically detectable impact on firm value except perhaps
in a handful of firms where the compensation plan is subject to a complete overhaul
Perhaps one explanation for the weak impact of SOP is that executive pay problems were
overstated or that by the time SOP was introduced (more than a decade after the Enron-type
scandals that led to calls for greater shareholder voice) they had been already addressed via
other mechanisms (hedge fund activism monitoring by institutional investors vote-no
campaigns against compensation committee members SEC-mandated pay disclosures)
Overall though based on the evidence to date it does not appear that SOP had a major impact14
Was the adoption of SOP an lsquooptimalrsquo choice from a social welfare point of view This is a
difficult question to answer given the challenges of identifying and measuring all the potential
costs and benefits associated with SOP (likely to change over time and differ across countries)
Perhaps the most positive view of SOP is that its introduction prevented the adoption of other
more radical and intrusive regulatory measures (eg CEO pay caps) In that sense it may be
viewed as an lsquooptimalrsquo answer (ie the lesser of two evils) to the political pressure to reform
executive pay during the financial crisis
14
This interpretation of the evidence is also consistent with the prediction of analytical studies Ozbas and
Matsusaka (2013) model the benefits and costs of shareholder empowerment distinguishing between the right to
approve and the right to propose They show that permitting shareholders to propose directors or policies can cause
value-maximizing managers to take value-reducing actions to accommodate activist investors with non-value-
maximizing goals As for the right to approve it is weakly beneficial that is approval rights (such as a SOP vote)
are generally beneficial for shareholders in that they limit the managerrsquos ability to pursue private benefits at
shareholder expense but they have minimal impact on managerial actions and firm value (because the manager in
effect can threaten shareholders with an undesirable status quo if they do not approve the managerrsquos proposed
action)
References
Associated Press 2007 Administration opposes say on pay bill April 19
Bainbridge S 2008 Remarks on say on pay An unjustified incursion on director authority
UCLA School of Law Research Paper No 08-06
Bebchuk L 2005 The Case for Increasing Shareholder Power Harvard Law Review 118833-
917
Bebchuk L 2007 Written testimony submitted before the Committee on Financial Services
United States House of Representatives Hearing on Empowering Shareholders on Executive
Compensation March 8
Bebchuk LA A Cohen and A Ferrell 2009 What Matters in Corporate Governance Review
of Financial Studies 22783-827
Cai J J Garner and R Walkling 2009 Electing Directors Journal of Finance 642387-2419
Cai J and R Walkling 2011 Shareholdersrsquo Say on Pay Does it Create Value Journal of
Financial and Quantitative Analysis 46299ndash339
Coates IV JC (2009) Testimony before the Subcommittee on Securities Insurance and
Investment of the Committee on Banking Housing and Urban Affairs United States Senate
July 29 2009 httppapersssrncomsol3paperscfmabstract_id=1475355
Correa R and U Lel 2013 Say On Pay Laws Executive Compensation CEO Pay Slice and
Firm Value Around the World Federal Reserve Board Working Paper
Cuntildeat V M Gine and M Guadalupe 2012 The Vote is Cast The Effect of Corporate
Governance on Shareholder Value Journal of Finance 671943-1977
Cuntildeat V M Gine and M Guadalupe 2013 Say Pays Shareholder Voice and Firm
Performance ECGI - Finance Working Paper No 373
Del Guercio D L Seery and T Woidtke 2008 Do Boards Pay Attention When Institutional
Investors lsquoJust Vote Norsquo Journal of Financial Economics 9084-103
Ertimur Y F Ferri and S Stubben 2010 Board of Directorslsquo Responsiveness to Shareholders
Evidence from Shareholder Proposals Journal of Corporate Finance 1653-72
Ertimur Y F Ferri and V Muslu 2011 Shareholder Activism and CEO Pay Review of
Financial Studies 24535ndash592
Ertimur Y F Ferri and D Oesch 2013 Shareholder Votes and Proxy Advisors Evidence from
Say on Pay Journal of Accounting Research 51951-996
Ertimur Y F Ferri and D Oesch forthcoming Does the Director Election System Matter
Evidence from Majority Voting Review of Accounting Studies forthcoming
Ferri F 2012 Low-Cost Shareholder Activism A Review of the Evidence Research
Handbook on the Economics of Corporate Law ed CA Hill and BH McDonnell
Edward Elgar Publishing
Ferri F and D Maber 2013 Say on Pay Votes and CEO Compensation Evidence from the UK
Review of Finance 17527-563
Ferri F and D Oesch 2013 Management Influence on Investors Evidence from Shareholder
Votes on the Frequency of Say on Pay Columbia Business School Research Paper No 13-17
Ferri F and T Sandino 2009 The impact of shareholder activism on financial reporting and
compensation The case of employee stock options expensing The Accounting Review 84433ndash
466
Ferri F and J Weber 2009 AFSCME vs Mozilohellipand Say on Pay for All (A) Harvard
Business School Case 109-009
Gompers P J Ishii and A Metrick 2003 Corporate Governance and Equity Prices Quarterly
Journal of Economics 118107-155
Gordon J 2009 Say on Paylsquo Cautionary Notes on the UK Experience and the Case for
Shareholders Opt-In Harvard Journal on Legislation 46323-368
Iliev P and S Vitanova 2013 The effect of the Say on Pay in the US Pennsylvania State
University Working Paper
Kaplan S 2007 Testimony of Steven N Kaplan on Empowering Shareholders on Executive
Compensation and HR 1257 the Shareholder Vote on Executive Compensation Act before the
Committee on Financial Services United States House of Representatives March 8
Larcker D A McCall and G Ormazabal 2013 The Economic Consequences of Proxy
Advisor Say-on-Pay Voting Policies Working Paper Stanford University
Larcker D G Ormazabal and D Taylor 2011 The market reaction to corporate governance
regulation Journal of Financial Economics 101431ndash448
Levit D and N Malenko 2011 Non-binding voting for shareholder proposals Journal of
Finance 661579ndash1614
Mishel L and N Sabadish 2012 How executive compensation and financial-sector pay have
fueled income inequality Issue Brief 331 Economic Policy Institute
Murphy K 2012 The Politics of Pay A Legislative History of Executive Compensation in
Jennifer Hill and Randall Thomas eds Research Handbook on Executive Pay Edward Elgar
Publishers
New York Times 2007 House votes to give investors say on executive pay April 21
Norris F 2004 Corporate democracy and the power to embarrass New York Times March 4
Ozbas O and J Matsusaka2013 Managerial Accommodation Proxy Access and the Cost of
Shareholder Empowerment University of Southern California CLEO Research Paper Series No
C12-1
Thomas R and J Cotter 2007 Shareholder proposals in the new millennium Shareholder
support board response and market reaction Journal of Corporate Finance 13 368ndash391
Thomas R A Palmiter and J Cotter 2012 Dodd-Franks Say on Pay Will it Lead to a Greater
Role for Shareholders in Corporate Governance Cornell Law Review 971213-1266
Thomas R and C Van der Elst 2013 The International Scope of Say on Pay ECGI Law
Working Paper No227
do not appear to be characterized by excess CEO pay However the authors also find significant
improvements in subsequent operating performance and efficiency as a result of passing SOP
proposals and thus conjecture that the SOP vote is viewed as a tool to express a vote of
confidence in management performance more than a way to change compensation practices and
that the large positive market reaction reflects expected performance improvements under this
tighter monitoring regime While this is an interesting idea it remains unclear what additional
ldquoteethrdquo a SOP vote may provide over existing (non-compensation related) monitoring
mechanisms shareholders can express (and do often express) their dissatisfaction with
management performance when voting on director elections (Cai Garner and Walkling 2009)
C Event studies around compensation changes induced by SOP
Two studies examine announcements of changes to compensation plans related to SOP Larcker
McCall and Ormazabal (2013) examine the market reaction to (non-contaminated) compensation
changes disclosed in 8-K filings by Russell 3000 firms in the US during the year before the first
SOP vote They find that changes that appear to be made to avoid a negative proxy advisorrsquos
recommendation and thus a negative SOP vote are associated with a negative abnormal return
of -044 whereas other compensation changes are not associated with a significant stock price
reaction Ertimur Ferri and Oesch (2013) perform a similar analysis but focusing on
compensation changes made after the first SOP vote (and explicitly in response to the vote) by
firms receiving a negative recommendation and a high voting dissent in 2011 They fail to find
significant abnormal returns even for the subset of compensation changes that resulted in a
positive recommendation and low dissent in 2012 (and thus were presumably perceived to be
adequate and material by proxy advisors and voting shareholders)
D Other approaches effect of SOP on Tobinrsquos Q
In their cross-country study Correa and Lel (2013) also examine the effect of SOP laws on
Tobinrsquos Q and report a 36 increase in firm value following the adoption of the SOP laws
(more precisely a 36 higher firm value for post-SOP observations relative to a control sample
that pools together pre-SOP and non-SOP observations) They acknowledge that this increase in
firm value is too large to be justified by the relative decrease in CEO pay that they document and
suggest (but do not test) that it may reflect better alignment of pay and performance13
A key
concern is that higher valuation in countries with SOP laws may reflect other governance
changes introduced at the same time While the study controls for other compensation-related
laws there may be other non-compensation regulations introduced with SOP and with a
potentially larger impact on firm value
IV What have we learned
Adoption of SOP in the US has been long advocated by those who contend that executive pay is
a manifestation of rather than a solution to the agency problem and by those who favor great
shareholder involvement in corporate decisions A growing body of research is starting to
investigate the effect of SOP on executive pay and firm value It is premature to draw definitive
conclusions also because these studies differ in methodologies and settings but three points
seem to emerge First there is robust evidence that boards respond to SOP votes when
13
They also find that the firm value increase is higher when the relative decrease in CEO pay results in a lower pay
differential between CEO and other top managers (CEO pay slice) implying that a source of value creation may be
the reduced pay inequality among the top management team But this seems unlikely to explain the change in
Tobinrsquos Q since the increase occurs for both firms with advisory SOP laws and firms with binding SOP laws and
there is no change in CEO pay slice in firms subject to binding SOP laws
shareholders use it ldquovoicerdquo is heardrdquo Both in the US and UK studies show that not only firms
failing to win the SOP vote but also firms facing substantial dissent (20-30 of the votes
against) make changes to their compensation plans in consultation with institutional investors
and proxy advisors Second in both the US and UK institutional investors appear to use the
power of SOP votes to pressure firms into strengthening the link (or the perceived link) between
pay and performance (particularly on the downside) but not to pressure firms to reduce target
levels of pay suggesting a reluctance of institutional investors to lsquoregulatersquo executive pay Third
and consistent with the above most studies examining the aggregate effect of SOP on CEO pay
find some increase in pay-performance sensitivity but not effect on the level and growth of CEO
pay While some observers may interpret the lack of an effect on CEO pay levels as evidence of
the failure of SOP one should note that SOP per se is a neutral tool Its impact depends on what
investors choose to lsquosay on payrsquo Fourth based on my interpretation of the existing research to
date there is little evidence of an effect of SOP on firm value The studies documenting a
positive price impact appear subject to alternative interpretations or not plausible in their
findings (eg they fail to find an effect on CEO pay or the effect is too small to justify the
documented price impact hence it remains unclear what the source of value creation is) Also
there seems to be no stock price reaction to the SOP-induced compensation changes Finally
given the nature of these SOP-induced changes while some of them may be beneficial it seems
hard to believe that they will have a statistically detectable impact on firm value except perhaps
in a handful of firms where the compensation plan is subject to a complete overhaul
Perhaps one explanation for the weak impact of SOP is that executive pay problems were
overstated or that by the time SOP was introduced (more than a decade after the Enron-type
scandals that led to calls for greater shareholder voice) they had been already addressed via
other mechanisms (hedge fund activism monitoring by institutional investors vote-no
campaigns against compensation committee members SEC-mandated pay disclosures)
Overall though based on the evidence to date it does not appear that SOP had a major impact14
Was the adoption of SOP an lsquooptimalrsquo choice from a social welfare point of view This is a
difficult question to answer given the challenges of identifying and measuring all the potential
costs and benefits associated with SOP (likely to change over time and differ across countries)
Perhaps the most positive view of SOP is that its introduction prevented the adoption of other
more radical and intrusive regulatory measures (eg CEO pay caps) In that sense it may be
viewed as an lsquooptimalrsquo answer (ie the lesser of two evils) to the political pressure to reform
executive pay during the financial crisis
14
This interpretation of the evidence is also consistent with the prediction of analytical studies Ozbas and
Matsusaka (2013) model the benefits and costs of shareholder empowerment distinguishing between the right to
approve and the right to propose They show that permitting shareholders to propose directors or policies can cause
value-maximizing managers to take value-reducing actions to accommodate activist investors with non-value-
maximizing goals As for the right to approve it is weakly beneficial that is approval rights (such as a SOP vote)
are generally beneficial for shareholders in that they limit the managerrsquos ability to pursue private benefits at
shareholder expense but they have minimal impact on managerial actions and firm value (because the manager in
effect can threaten shareholders with an undesirable status quo if they do not approve the managerrsquos proposed
action)
References
Associated Press 2007 Administration opposes say on pay bill April 19
Bainbridge S 2008 Remarks on say on pay An unjustified incursion on director authority
UCLA School of Law Research Paper No 08-06
Bebchuk L 2005 The Case for Increasing Shareholder Power Harvard Law Review 118833-
917
Bebchuk L 2007 Written testimony submitted before the Committee on Financial Services
United States House of Representatives Hearing on Empowering Shareholders on Executive
Compensation March 8
Bebchuk LA A Cohen and A Ferrell 2009 What Matters in Corporate Governance Review
of Financial Studies 22783-827
Cai J J Garner and R Walkling 2009 Electing Directors Journal of Finance 642387-2419
Cai J and R Walkling 2011 Shareholdersrsquo Say on Pay Does it Create Value Journal of
Financial and Quantitative Analysis 46299ndash339
Coates IV JC (2009) Testimony before the Subcommittee on Securities Insurance and
Investment of the Committee on Banking Housing and Urban Affairs United States Senate
July 29 2009 httppapersssrncomsol3paperscfmabstract_id=1475355
Correa R and U Lel 2013 Say On Pay Laws Executive Compensation CEO Pay Slice and
Firm Value Around the World Federal Reserve Board Working Paper
Cuntildeat V M Gine and M Guadalupe 2012 The Vote is Cast The Effect of Corporate
Governance on Shareholder Value Journal of Finance 671943-1977
Cuntildeat V M Gine and M Guadalupe 2013 Say Pays Shareholder Voice and Firm
Performance ECGI - Finance Working Paper No 373
Del Guercio D L Seery and T Woidtke 2008 Do Boards Pay Attention When Institutional
Investors lsquoJust Vote Norsquo Journal of Financial Economics 9084-103
Ertimur Y F Ferri and S Stubben 2010 Board of Directorslsquo Responsiveness to Shareholders
Evidence from Shareholder Proposals Journal of Corporate Finance 1653-72
Ertimur Y F Ferri and V Muslu 2011 Shareholder Activism and CEO Pay Review of
Financial Studies 24535ndash592
Ertimur Y F Ferri and D Oesch 2013 Shareholder Votes and Proxy Advisors Evidence from
Say on Pay Journal of Accounting Research 51951-996
Ertimur Y F Ferri and D Oesch forthcoming Does the Director Election System Matter
Evidence from Majority Voting Review of Accounting Studies forthcoming
Ferri F 2012 Low-Cost Shareholder Activism A Review of the Evidence Research
Handbook on the Economics of Corporate Law ed CA Hill and BH McDonnell
Edward Elgar Publishing
Ferri F and D Maber 2013 Say on Pay Votes and CEO Compensation Evidence from the UK
Review of Finance 17527-563
Ferri F and D Oesch 2013 Management Influence on Investors Evidence from Shareholder
Votes on the Frequency of Say on Pay Columbia Business School Research Paper No 13-17
Ferri F and T Sandino 2009 The impact of shareholder activism on financial reporting and
compensation The case of employee stock options expensing The Accounting Review 84433ndash
466
Ferri F and J Weber 2009 AFSCME vs Mozilohellipand Say on Pay for All (A) Harvard
Business School Case 109-009
Gompers P J Ishii and A Metrick 2003 Corporate Governance and Equity Prices Quarterly
Journal of Economics 118107-155
Gordon J 2009 Say on Paylsquo Cautionary Notes on the UK Experience and the Case for
Shareholders Opt-In Harvard Journal on Legislation 46323-368
Iliev P and S Vitanova 2013 The effect of the Say on Pay in the US Pennsylvania State
University Working Paper
Kaplan S 2007 Testimony of Steven N Kaplan on Empowering Shareholders on Executive
Compensation and HR 1257 the Shareholder Vote on Executive Compensation Act before the
Committee on Financial Services United States House of Representatives March 8
Larcker D A McCall and G Ormazabal 2013 The Economic Consequences of Proxy
Advisor Say-on-Pay Voting Policies Working Paper Stanford University
Larcker D G Ormazabal and D Taylor 2011 The market reaction to corporate governance
regulation Journal of Financial Economics 101431ndash448
Levit D and N Malenko 2011 Non-binding voting for shareholder proposals Journal of
Finance 661579ndash1614
Mishel L and N Sabadish 2012 How executive compensation and financial-sector pay have
fueled income inequality Issue Brief 331 Economic Policy Institute
Murphy K 2012 The Politics of Pay A Legislative History of Executive Compensation in
Jennifer Hill and Randall Thomas eds Research Handbook on Executive Pay Edward Elgar
Publishers
New York Times 2007 House votes to give investors say on executive pay April 21
Norris F 2004 Corporate democracy and the power to embarrass New York Times March 4
Ozbas O and J Matsusaka2013 Managerial Accommodation Proxy Access and the Cost of
Shareholder Empowerment University of Southern California CLEO Research Paper Series No
C12-1
Thomas R and J Cotter 2007 Shareholder proposals in the new millennium Shareholder
support board response and market reaction Journal of Corporate Finance 13 368ndash391
Thomas R A Palmiter and J Cotter 2012 Dodd-Franks Say on Pay Will it Lead to a Greater
Role for Shareholders in Corporate Governance Cornell Law Review 971213-1266
Thomas R and C Van der Elst 2013 The International Scope of Say on Pay ECGI Law
Working Paper No227
D Other approaches effect of SOP on Tobinrsquos Q
In their cross-country study Correa and Lel (2013) also examine the effect of SOP laws on
Tobinrsquos Q and report a 36 increase in firm value following the adoption of the SOP laws
(more precisely a 36 higher firm value for post-SOP observations relative to a control sample
that pools together pre-SOP and non-SOP observations) They acknowledge that this increase in
firm value is too large to be justified by the relative decrease in CEO pay that they document and
suggest (but do not test) that it may reflect better alignment of pay and performance13
A key
concern is that higher valuation in countries with SOP laws may reflect other governance
changes introduced at the same time While the study controls for other compensation-related
laws there may be other non-compensation regulations introduced with SOP and with a
potentially larger impact on firm value
IV What have we learned
Adoption of SOP in the US has been long advocated by those who contend that executive pay is
a manifestation of rather than a solution to the agency problem and by those who favor great
shareholder involvement in corporate decisions A growing body of research is starting to
investigate the effect of SOP on executive pay and firm value It is premature to draw definitive
conclusions also because these studies differ in methodologies and settings but three points
seem to emerge First there is robust evidence that boards respond to SOP votes when
13
They also find that the firm value increase is higher when the relative decrease in CEO pay results in a lower pay
differential between CEO and other top managers (CEO pay slice) implying that a source of value creation may be
the reduced pay inequality among the top management team But this seems unlikely to explain the change in
Tobinrsquos Q since the increase occurs for both firms with advisory SOP laws and firms with binding SOP laws and
there is no change in CEO pay slice in firms subject to binding SOP laws
shareholders use it ldquovoicerdquo is heardrdquo Both in the US and UK studies show that not only firms
failing to win the SOP vote but also firms facing substantial dissent (20-30 of the votes
against) make changes to their compensation plans in consultation with institutional investors
and proxy advisors Second in both the US and UK institutional investors appear to use the
power of SOP votes to pressure firms into strengthening the link (or the perceived link) between
pay and performance (particularly on the downside) but not to pressure firms to reduce target
levels of pay suggesting a reluctance of institutional investors to lsquoregulatersquo executive pay Third
and consistent with the above most studies examining the aggregate effect of SOP on CEO pay
find some increase in pay-performance sensitivity but not effect on the level and growth of CEO
pay While some observers may interpret the lack of an effect on CEO pay levels as evidence of
the failure of SOP one should note that SOP per se is a neutral tool Its impact depends on what
investors choose to lsquosay on payrsquo Fourth based on my interpretation of the existing research to
date there is little evidence of an effect of SOP on firm value The studies documenting a
positive price impact appear subject to alternative interpretations or not plausible in their
findings (eg they fail to find an effect on CEO pay or the effect is too small to justify the
documented price impact hence it remains unclear what the source of value creation is) Also
there seems to be no stock price reaction to the SOP-induced compensation changes Finally
given the nature of these SOP-induced changes while some of them may be beneficial it seems
hard to believe that they will have a statistically detectable impact on firm value except perhaps
in a handful of firms where the compensation plan is subject to a complete overhaul
Perhaps one explanation for the weak impact of SOP is that executive pay problems were
overstated or that by the time SOP was introduced (more than a decade after the Enron-type
scandals that led to calls for greater shareholder voice) they had been already addressed via
other mechanisms (hedge fund activism monitoring by institutional investors vote-no
campaigns against compensation committee members SEC-mandated pay disclosures)
Overall though based on the evidence to date it does not appear that SOP had a major impact14
Was the adoption of SOP an lsquooptimalrsquo choice from a social welfare point of view This is a
difficult question to answer given the challenges of identifying and measuring all the potential
costs and benefits associated with SOP (likely to change over time and differ across countries)
Perhaps the most positive view of SOP is that its introduction prevented the adoption of other
more radical and intrusive regulatory measures (eg CEO pay caps) In that sense it may be
viewed as an lsquooptimalrsquo answer (ie the lesser of two evils) to the political pressure to reform
executive pay during the financial crisis
14
This interpretation of the evidence is also consistent with the prediction of analytical studies Ozbas and
Matsusaka (2013) model the benefits and costs of shareholder empowerment distinguishing between the right to
approve and the right to propose They show that permitting shareholders to propose directors or policies can cause
value-maximizing managers to take value-reducing actions to accommodate activist investors with non-value-
maximizing goals As for the right to approve it is weakly beneficial that is approval rights (such as a SOP vote)
are generally beneficial for shareholders in that they limit the managerrsquos ability to pursue private benefits at
shareholder expense but they have minimal impact on managerial actions and firm value (because the manager in
effect can threaten shareholders with an undesirable status quo if they do not approve the managerrsquos proposed
action)
References
Associated Press 2007 Administration opposes say on pay bill April 19
Bainbridge S 2008 Remarks on say on pay An unjustified incursion on director authority
UCLA School of Law Research Paper No 08-06
Bebchuk L 2005 The Case for Increasing Shareholder Power Harvard Law Review 118833-
917
Bebchuk L 2007 Written testimony submitted before the Committee on Financial Services
United States House of Representatives Hearing on Empowering Shareholders on Executive
Compensation March 8
Bebchuk LA A Cohen and A Ferrell 2009 What Matters in Corporate Governance Review
of Financial Studies 22783-827
Cai J J Garner and R Walkling 2009 Electing Directors Journal of Finance 642387-2419
Cai J and R Walkling 2011 Shareholdersrsquo Say on Pay Does it Create Value Journal of
Financial and Quantitative Analysis 46299ndash339
Coates IV JC (2009) Testimony before the Subcommittee on Securities Insurance and
Investment of the Committee on Banking Housing and Urban Affairs United States Senate
July 29 2009 httppapersssrncomsol3paperscfmabstract_id=1475355
Correa R and U Lel 2013 Say On Pay Laws Executive Compensation CEO Pay Slice and
Firm Value Around the World Federal Reserve Board Working Paper
Cuntildeat V M Gine and M Guadalupe 2012 The Vote is Cast The Effect of Corporate
Governance on Shareholder Value Journal of Finance 671943-1977
Cuntildeat V M Gine and M Guadalupe 2013 Say Pays Shareholder Voice and Firm
Performance ECGI - Finance Working Paper No 373
Del Guercio D L Seery and T Woidtke 2008 Do Boards Pay Attention When Institutional
Investors lsquoJust Vote Norsquo Journal of Financial Economics 9084-103
Ertimur Y F Ferri and S Stubben 2010 Board of Directorslsquo Responsiveness to Shareholders
Evidence from Shareholder Proposals Journal of Corporate Finance 1653-72
Ertimur Y F Ferri and V Muslu 2011 Shareholder Activism and CEO Pay Review of
Financial Studies 24535ndash592
Ertimur Y F Ferri and D Oesch 2013 Shareholder Votes and Proxy Advisors Evidence from
Say on Pay Journal of Accounting Research 51951-996
Ertimur Y F Ferri and D Oesch forthcoming Does the Director Election System Matter
Evidence from Majority Voting Review of Accounting Studies forthcoming
Ferri F 2012 Low-Cost Shareholder Activism A Review of the Evidence Research
Handbook on the Economics of Corporate Law ed CA Hill and BH McDonnell
Edward Elgar Publishing
Ferri F and D Maber 2013 Say on Pay Votes and CEO Compensation Evidence from the UK
Review of Finance 17527-563
Ferri F and D Oesch 2013 Management Influence on Investors Evidence from Shareholder
Votes on the Frequency of Say on Pay Columbia Business School Research Paper No 13-17
Ferri F and T Sandino 2009 The impact of shareholder activism on financial reporting and
compensation The case of employee stock options expensing The Accounting Review 84433ndash
466
Ferri F and J Weber 2009 AFSCME vs Mozilohellipand Say on Pay for All (A) Harvard
Business School Case 109-009
Gompers P J Ishii and A Metrick 2003 Corporate Governance and Equity Prices Quarterly
Journal of Economics 118107-155
Gordon J 2009 Say on Paylsquo Cautionary Notes on the UK Experience and the Case for
Shareholders Opt-In Harvard Journal on Legislation 46323-368
Iliev P and S Vitanova 2013 The effect of the Say on Pay in the US Pennsylvania State
University Working Paper
Kaplan S 2007 Testimony of Steven N Kaplan on Empowering Shareholders on Executive
Compensation and HR 1257 the Shareholder Vote on Executive Compensation Act before the
Committee on Financial Services United States House of Representatives March 8
Larcker D A McCall and G Ormazabal 2013 The Economic Consequences of Proxy
Advisor Say-on-Pay Voting Policies Working Paper Stanford University
Larcker D G Ormazabal and D Taylor 2011 The market reaction to corporate governance
regulation Journal of Financial Economics 101431ndash448
Levit D and N Malenko 2011 Non-binding voting for shareholder proposals Journal of
Finance 661579ndash1614
Mishel L and N Sabadish 2012 How executive compensation and financial-sector pay have
fueled income inequality Issue Brief 331 Economic Policy Institute
Murphy K 2012 The Politics of Pay A Legislative History of Executive Compensation in
Jennifer Hill and Randall Thomas eds Research Handbook on Executive Pay Edward Elgar
Publishers
New York Times 2007 House votes to give investors say on executive pay April 21
Norris F 2004 Corporate democracy and the power to embarrass New York Times March 4
Ozbas O and J Matsusaka2013 Managerial Accommodation Proxy Access and the Cost of
Shareholder Empowerment University of Southern California CLEO Research Paper Series No
C12-1
Thomas R and J Cotter 2007 Shareholder proposals in the new millennium Shareholder
support board response and market reaction Journal of Corporate Finance 13 368ndash391
Thomas R A Palmiter and J Cotter 2012 Dodd-Franks Say on Pay Will it Lead to a Greater
Role for Shareholders in Corporate Governance Cornell Law Review 971213-1266
Thomas R and C Van der Elst 2013 The International Scope of Say on Pay ECGI Law
Working Paper No227
shareholders use it ldquovoicerdquo is heardrdquo Both in the US and UK studies show that not only firms
failing to win the SOP vote but also firms facing substantial dissent (20-30 of the votes
against) make changes to their compensation plans in consultation with institutional investors
and proxy advisors Second in both the US and UK institutional investors appear to use the
power of SOP votes to pressure firms into strengthening the link (or the perceived link) between
pay and performance (particularly on the downside) but not to pressure firms to reduce target
levels of pay suggesting a reluctance of institutional investors to lsquoregulatersquo executive pay Third
and consistent with the above most studies examining the aggregate effect of SOP on CEO pay
find some increase in pay-performance sensitivity but not effect on the level and growth of CEO
pay While some observers may interpret the lack of an effect on CEO pay levels as evidence of
the failure of SOP one should note that SOP per se is a neutral tool Its impact depends on what
investors choose to lsquosay on payrsquo Fourth based on my interpretation of the existing research to
date there is little evidence of an effect of SOP on firm value The studies documenting a
positive price impact appear subject to alternative interpretations or not plausible in their
findings (eg they fail to find an effect on CEO pay or the effect is too small to justify the
documented price impact hence it remains unclear what the source of value creation is) Also
there seems to be no stock price reaction to the SOP-induced compensation changes Finally
given the nature of these SOP-induced changes while some of them may be beneficial it seems
hard to believe that they will have a statistically detectable impact on firm value except perhaps
in a handful of firms where the compensation plan is subject to a complete overhaul
Perhaps one explanation for the weak impact of SOP is that executive pay problems were
overstated or that by the time SOP was introduced (more than a decade after the Enron-type
scandals that led to calls for greater shareholder voice) they had been already addressed via
other mechanisms (hedge fund activism monitoring by institutional investors vote-no
campaigns against compensation committee members SEC-mandated pay disclosures)
Overall though based on the evidence to date it does not appear that SOP had a major impact14
Was the adoption of SOP an lsquooptimalrsquo choice from a social welfare point of view This is a
difficult question to answer given the challenges of identifying and measuring all the potential
costs and benefits associated with SOP (likely to change over time and differ across countries)
Perhaps the most positive view of SOP is that its introduction prevented the adoption of other
more radical and intrusive regulatory measures (eg CEO pay caps) In that sense it may be
viewed as an lsquooptimalrsquo answer (ie the lesser of two evils) to the political pressure to reform
executive pay during the financial crisis
14
This interpretation of the evidence is also consistent with the prediction of analytical studies Ozbas and
Matsusaka (2013) model the benefits and costs of shareholder empowerment distinguishing between the right to
approve and the right to propose They show that permitting shareholders to propose directors or policies can cause
value-maximizing managers to take value-reducing actions to accommodate activist investors with non-value-
maximizing goals As for the right to approve it is weakly beneficial that is approval rights (such as a SOP vote)
are generally beneficial for shareholders in that they limit the managerrsquos ability to pursue private benefits at
shareholder expense but they have minimal impact on managerial actions and firm value (because the manager in
effect can threaten shareholders with an undesirable status quo if they do not approve the managerrsquos proposed
action)
References
Associated Press 2007 Administration opposes say on pay bill April 19
Bainbridge S 2008 Remarks on say on pay An unjustified incursion on director authority
UCLA School of Law Research Paper No 08-06
Bebchuk L 2005 The Case for Increasing Shareholder Power Harvard Law Review 118833-
917
Bebchuk L 2007 Written testimony submitted before the Committee on Financial Services
United States House of Representatives Hearing on Empowering Shareholders on Executive
Compensation March 8
Bebchuk LA A Cohen and A Ferrell 2009 What Matters in Corporate Governance Review
of Financial Studies 22783-827
Cai J J Garner and R Walkling 2009 Electing Directors Journal of Finance 642387-2419
Cai J and R Walkling 2011 Shareholdersrsquo Say on Pay Does it Create Value Journal of
Financial and Quantitative Analysis 46299ndash339
Coates IV JC (2009) Testimony before the Subcommittee on Securities Insurance and
Investment of the Committee on Banking Housing and Urban Affairs United States Senate
July 29 2009 httppapersssrncomsol3paperscfmabstract_id=1475355
Correa R and U Lel 2013 Say On Pay Laws Executive Compensation CEO Pay Slice and
Firm Value Around the World Federal Reserve Board Working Paper
Cuntildeat V M Gine and M Guadalupe 2012 The Vote is Cast The Effect of Corporate
Governance on Shareholder Value Journal of Finance 671943-1977
Cuntildeat V M Gine and M Guadalupe 2013 Say Pays Shareholder Voice and Firm
Performance ECGI - Finance Working Paper No 373
Del Guercio D L Seery and T Woidtke 2008 Do Boards Pay Attention When Institutional
Investors lsquoJust Vote Norsquo Journal of Financial Economics 9084-103
Ertimur Y F Ferri and S Stubben 2010 Board of Directorslsquo Responsiveness to Shareholders
Evidence from Shareholder Proposals Journal of Corporate Finance 1653-72
Ertimur Y F Ferri and V Muslu 2011 Shareholder Activism and CEO Pay Review of
Financial Studies 24535ndash592
Ertimur Y F Ferri and D Oesch 2013 Shareholder Votes and Proxy Advisors Evidence from
Say on Pay Journal of Accounting Research 51951-996
Ertimur Y F Ferri and D Oesch forthcoming Does the Director Election System Matter
Evidence from Majority Voting Review of Accounting Studies forthcoming
Ferri F 2012 Low-Cost Shareholder Activism A Review of the Evidence Research
Handbook on the Economics of Corporate Law ed CA Hill and BH McDonnell
Edward Elgar Publishing
Ferri F and D Maber 2013 Say on Pay Votes and CEO Compensation Evidence from the UK
Review of Finance 17527-563
Ferri F and D Oesch 2013 Management Influence on Investors Evidence from Shareholder
Votes on the Frequency of Say on Pay Columbia Business School Research Paper No 13-17
Ferri F and T Sandino 2009 The impact of shareholder activism on financial reporting and
compensation The case of employee stock options expensing The Accounting Review 84433ndash
466
Ferri F and J Weber 2009 AFSCME vs Mozilohellipand Say on Pay for All (A) Harvard
Business School Case 109-009
Gompers P J Ishii and A Metrick 2003 Corporate Governance and Equity Prices Quarterly
Journal of Economics 118107-155
Gordon J 2009 Say on Paylsquo Cautionary Notes on the UK Experience and the Case for
Shareholders Opt-In Harvard Journal on Legislation 46323-368
Iliev P and S Vitanova 2013 The effect of the Say on Pay in the US Pennsylvania State
University Working Paper
Kaplan S 2007 Testimony of Steven N Kaplan on Empowering Shareholders on Executive
Compensation and HR 1257 the Shareholder Vote on Executive Compensation Act before the
Committee on Financial Services United States House of Representatives March 8
Larcker D A McCall and G Ormazabal 2013 The Economic Consequences of Proxy
Advisor Say-on-Pay Voting Policies Working Paper Stanford University
Larcker D G Ormazabal and D Taylor 2011 The market reaction to corporate governance
regulation Journal of Financial Economics 101431ndash448
Levit D and N Malenko 2011 Non-binding voting for shareholder proposals Journal of
Finance 661579ndash1614
Mishel L and N Sabadish 2012 How executive compensation and financial-sector pay have
fueled income inequality Issue Brief 331 Economic Policy Institute
Murphy K 2012 The Politics of Pay A Legislative History of Executive Compensation in
Jennifer Hill and Randall Thomas eds Research Handbook on Executive Pay Edward Elgar
Publishers
New York Times 2007 House votes to give investors say on executive pay April 21
Norris F 2004 Corporate democracy and the power to embarrass New York Times March 4
Ozbas O and J Matsusaka2013 Managerial Accommodation Proxy Access and the Cost of
Shareholder Empowerment University of Southern California CLEO Research Paper Series No
C12-1
Thomas R and J Cotter 2007 Shareholder proposals in the new millennium Shareholder
support board response and market reaction Journal of Corporate Finance 13 368ndash391
Thomas R A Palmiter and J Cotter 2012 Dodd-Franks Say on Pay Will it Lead to a Greater
Role for Shareholders in Corporate Governance Cornell Law Review 971213-1266
Thomas R and C Van der Elst 2013 The International Scope of Say on Pay ECGI Law
Working Paper No227
scandals that led to calls for greater shareholder voice) they had been already addressed via
other mechanisms (hedge fund activism monitoring by institutional investors vote-no
campaigns against compensation committee members SEC-mandated pay disclosures)
Overall though based on the evidence to date it does not appear that SOP had a major impact14
Was the adoption of SOP an lsquooptimalrsquo choice from a social welfare point of view This is a
difficult question to answer given the challenges of identifying and measuring all the potential
costs and benefits associated with SOP (likely to change over time and differ across countries)
Perhaps the most positive view of SOP is that its introduction prevented the adoption of other
more radical and intrusive regulatory measures (eg CEO pay caps) In that sense it may be
viewed as an lsquooptimalrsquo answer (ie the lesser of two evils) to the political pressure to reform
executive pay during the financial crisis
14
This interpretation of the evidence is also consistent with the prediction of analytical studies Ozbas and
Matsusaka (2013) model the benefits and costs of shareholder empowerment distinguishing between the right to
approve and the right to propose They show that permitting shareholders to propose directors or policies can cause
value-maximizing managers to take value-reducing actions to accommodate activist investors with non-value-
maximizing goals As for the right to approve it is weakly beneficial that is approval rights (such as a SOP vote)
are generally beneficial for shareholders in that they limit the managerrsquos ability to pursue private benefits at
shareholder expense but they have minimal impact on managerial actions and firm value (because the manager in
effect can threaten shareholders with an undesirable status quo if they do not approve the managerrsquos proposed
action)
References
Associated Press 2007 Administration opposes say on pay bill April 19
Bainbridge S 2008 Remarks on say on pay An unjustified incursion on director authority
UCLA School of Law Research Paper No 08-06
Bebchuk L 2005 The Case for Increasing Shareholder Power Harvard Law Review 118833-
917
Bebchuk L 2007 Written testimony submitted before the Committee on Financial Services
United States House of Representatives Hearing on Empowering Shareholders on Executive
Compensation March 8
Bebchuk LA A Cohen and A Ferrell 2009 What Matters in Corporate Governance Review
of Financial Studies 22783-827
Cai J J Garner and R Walkling 2009 Electing Directors Journal of Finance 642387-2419
Cai J and R Walkling 2011 Shareholdersrsquo Say on Pay Does it Create Value Journal of
Financial and Quantitative Analysis 46299ndash339
Coates IV JC (2009) Testimony before the Subcommittee on Securities Insurance and
Investment of the Committee on Banking Housing and Urban Affairs United States Senate
July 29 2009 httppapersssrncomsol3paperscfmabstract_id=1475355
Correa R and U Lel 2013 Say On Pay Laws Executive Compensation CEO Pay Slice and
Firm Value Around the World Federal Reserve Board Working Paper
Cuntildeat V M Gine and M Guadalupe 2012 The Vote is Cast The Effect of Corporate
Governance on Shareholder Value Journal of Finance 671943-1977
Cuntildeat V M Gine and M Guadalupe 2013 Say Pays Shareholder Voice and Firm
Performance ECGI - Finance Working Paper No 373
Del Guercio D L Seery and T Woidtke 2008 Do Boards Pay Attention When Institutional
Investors lsquoJust Vote Norsquo Journal of Financial Economics 9084-103
Ertimur Y F Ferri and S Stubben 2010 Board of Directorslsquo Responsiveness to Shareholders
Evidence from Shareholder Proposals Journal of Corporate Finance 1653-72
Ertimur Y F Ferri and V Muslu 2011 Shareholder Activism and CEO Pay Review of
Financial Studies 24535ndash592
Ertimur Y F Ferri and D Oesch 2013 Shareholder Votes and Proxy Advisors Evidence from
Say on Pay Journal of Accounting Research 51951-996
Ertimur Y F Ferri and D Oesch forthcoming Does the Director Election System Matter
Evidence from Majority Voting Review of Accounting Studies forthcoming
Ferri F 2012 Low-Cost Shareholder Activism A Review of the Evidence Research
Handbook on the Economics of Corporate Law ed CA Hill and BH McDonnell
Edward Elgar Publishing
Ferri F and D Maber 2013 Say on Pay Votes and CEO Compensation Evidence from the UK
Review of Finance 17527-563
Ferri F and D Oesch 2013 Management Influence on Investors Evidence from Shareholder
Votes on the Frequency of Say on Pay Columbia Business School Research Paper No 13-17
Ferri F and T Sandino 2009 The impact of shareholder activism on financial reporting and
compensation The case of employee stock options expensing The Accounting Review 84433ndash
466
Ferri F and J Weber 2009 AFSCME vs Mozilohellipand Say on Pay for All (A) Harvard
Business School Case 109-009
Gompers P J Ishii and A Metrick 2003 Corporate Governance and Equity Prices Quarterly
Journal of Economics 118107-155
Gordon J 2009 Say on Paylsquo Cautionary Notes on the UK Experience and the Case for
Shareholders Opt-In Harvard Journal on Legislation 46323-368
Iliev P and S Vitanova 2013 The effect of the Say on Pay in the US Pennsylvania State
University Working Paper
Kaplan S 2007 Testimony of Steven N Kaplan on Empowering Shareholders on Executive
Compensation and HR 1257 the Shareholder Vote on Executive Compensation Act before the
Committee on Financial Services United States House of Representatives March 8
Larcker D A McCall and G Ormazabal 2013 The Economic Consequences of Proxy
Advisor Say-on-Pay Voting Policies Working Paper Stanford University
Larcker D G Ormazabal and D Taylor 2011 The market reaction to corporate governance
regulation Journal of Financial Economics 101431ndash448
Levit D and N Malenko 2011 Non-binding voting for shareholder proposals Journal of
Finance 661579ndash1614
Mishel L and N Sabadish 2012 How executive compensation and financial-sector pay have
fueled income inequality Issue Brief 331 Economic Policy Institute
Murphy K 2012 The Politics of Pay A Legislative History of Executive Compensation in
Jennifer Hill and Randall Thomas eds Research Handbook on Executive Pay Edward Elgar
Publishers
New York Times 2007 House votes to give investors say on executive pay April 21
Norris F 2004 Corporate democracy and the power to embarrass New York Times March 4
Ozbas O and J Matsusaka2013 Managerial Accommodation Proxy Access and the Cost of
Shareholder Empowerment University of Southern California CLEO Research Paper Series No
C12-1
Thomas R and J Cotter 2007 Shareholder proposals in the new millennium Shareholder
support board response and market reaction Journal of Corporate Finance 13 368ndash391
Thomas R A Palmiter and J Cotter 2012 Dodd-Franks Say on Pay Will it Lead to a Greater
Role for Shareholders in Corporate Governance Cornell Law Review 971213-1266
Thomas R and C Van der Elst 2013 The International Scope of Say on Pay ECGI Law
Working Paper No227
References
Associated Press 2007 Administration opposes say on pay bill April 19
Bainbridge S 2008 Remarks on say on pay An unjustified incursion on director authority
UCLA School of Law Research Paper No 08-06
Bebchuk L 2005 The Case for Increasing Shareholder Power Harvard Law Review 118833-
917
Bebchuk L 2007 Written testimony submitted before the Committee on Financial Services
United States House of Representatives Hearing on Empowering Shareholders on Executive
Compensation March 8
Bebchuk LA A Cohen and A Ferrell 2009 What Matters in Corporate Governance Review
of Financial Studies 22783-827
Cai J J Garner and R Walkling 2009 Electing Directors Journal of Finance 642387-2419
Cai J and R Walkling 2011 Shareholdersrsquo Say on Pay Does it Create Value Journal of
Financial and Quantitative Analysis 46299ndash339
Coates IV JC (2009) Testimony before the Subcommittee on Securities Insurance and
Investment of the Committee on Banking Housing and Urban Affairs United States Senate
July 29 2009 httppapersssrncomsol3paperscfmabstract_id=1475355
Correa R and U Lel 2013 Say On Pay Laws Executive Compensation CEO Pay Slice and
Firm Value Around the World Federal Reserve Board Working Paper
Cuntildeat V M Gine and M Guadalupe 2012 The Vote is Cast The Effect of Corporate
Governance on Shareholder Value Journal of Finance 671943-1977
Cuntildeat V M Gine and M Guadalupe 2013 Say Pays Shareholder Voice and Firm
Performance ECGI - Finance Working Paper No 373
Del Guercio D L Seery and T Woidtke 2008 Do Boards Pay Attention When Institutional
Investors lsquoJust Vote Norsquo Journal of Financial Economics 9084-103
Ertimur Y F Ferri and S Stubben 2010 Board of Directorslsquo Responsiveness to Shareholders
Evidence from Shareholder Proposals Journal of Corporate Finance 1653-72
Ertimur Y F Ferri and V Muslu 2011 Shareholder Activism and CEO Pay Review of
Financial Studies 24535ndash592
Ertimur Y F Ferri and D Oesch 2013 Shareholder Votes and Proxy Advisors Evidence from
Say on Pay Journal of Accounting Research 51951-996
Ertimur Y F Ferri and D Oesch forthcoming Does the Director Election System Matter
Evidence from Majority Voting Review of Accounting Studies forthcoming
Ferri F 2012 Low-Cost Shareholder Activism A Review of the Evidence Research
Handbook on the Economics of Corporate Law ed CA Hill and BH McDonnell
Edward Elgar Publishing
Ferri F and D Maber 2013 Say on Pay Votes and CEO Compensation Evidence from the UK
Review of Finance 17527-563
Ferri F and D Oesch 2013 Management Influence on Investors Evidence from Shareholder
Votes on the Frequency of Say on Pay Columbia Business School Research Paper No 13-17
Ferri F and T Sandino 2009 The impact of shareholder activism on financial reporting and
compensation The case of employee stock options expensing The Accounting Review 84433ndash
466
Ferri F and J Weber 2009 AFSCME vs Mozilohellipand Say on Pay for All (A) Harvard
Business School Case 109-009
Gompers P J Ishii and A Metrick 2003 Corporate Governance and Equity Prices Quarterly
Journal of Economics 118107-155
Gordon J 2009 Say on Paylsquo Cautionary Notes on the UK Experience and the Case for
Shareholders Opt-In Harvard Journal on Legislation 46323-368
Iliev P and S Vitanova 2013 The effect of the Say on Pay in the US Pennsylvania State
University Working Paper
Kaplan S 2007 Testimony of Steven N Kaplan on Empowering Shareholders on Executive
Compensation and HR 1257 the Shareholder Vote on Executive Compensation Act before the
Committee on Financial Services United States House of Representatives March 8
Larcker D A McCall and G Ormazabal 2013 The Economic Consequences of Proxy
Advisor Say-on-Pay Voting Policies Working Paper Stanford University
Larcker D G Ormazabal and D Taylor 2011 The market reaction to corporate governance
regulation Journal of Financial Economics 101431ndash448
Levit D and N Malenko 2011 Non-binding voting for shareholder proposals Journal of
Finance 661579ndash1614
Mishel L and N Sabadish 2012 How executive compensation and financial-sector pay have
fueled income inequality Issue Brief 331 Economic Policy Institute
Murphy K 2012 The Politics of Pay A Legislative History of Executive Compensation in
Jennifer Hill and Randall Thomas eds Research Handbook on Executive Pay Edward Elgar
Publishers
New York Times 2007 House votes to give investors say on executive pay April 21
Norris F 2004 Corporate democracy and the power to embarrass New York Times March 4
Ozbas O and J Matsusaka2013 Managerial Accommodation Proxy Access and the Cost of
Shareholder Empowerment University of Southern California CLEO Research Paper Series No
C12-1
Thomas R and J Cotter 2007 Shareholder proposals in the new millennium Shareholder
support board response and market reaction Journal of Corporate Finance 13 368ndash391
Thomas R A Palmiter and J Cotter 2012 Dodd-Franks Say on Pay Will it Lead to a Greater
Role for Shareholders in Corporate Governance Cornell Law Review 971213-1266
Thomas R and C Van der Elst 2013 The International Scope of Say on Pay ECGI Law
Working Paper No227
Cuntildeat V M Gine and M Guadalupe 2013 Say Pays Shareholder Voice and Firm
Performance ECGI - Finance Working Paper No 373
Del Guercio D L Seery and T Woidtke 2008 Do Boards Pay Attention When Institutional
Investors lsquoJust Vote Norsquo Journal of Financial Economics 9084-103
Ertimur Y F Ferri and S Stubben 2010 Board of Directorslsquo Responsiveness to Shareholders
Evidence from Shareholder Proposals Journal of Corporate Finance 1653-72
Ertimur Y F Ferri and V Muslu 2011 Shareholder Activism and CEO Pay Review of
Financial Studies 24535ndash592
Ertimur Y F Ferri and D Oesch 2013 Shareholder Votes and Proxy Advisors Evidence from
Say on Pay Journal of Accounting Research 51951-996
Ertimur Y F Ferri and D Oesch forthcoming Does the Director Election System Matter
Evidence from Majority Voting Review of Accounting Studies forthcoming
Ferri F 2012 Low-Cost Shareholder Activism A Review of the Evidence Research
Handbook on the Economics of Corporate Law ed CA Hill and BH McDonnell
Edward Elgar Publishing
Ferri F and D Maber 2013 Say on Pay Votes and CEO Compensation Evidence from the UK
Review of Finance 17527-563
Ferri F and D Oesch 2013 Management Influence on Investors Evidence from Shareholder
Votes on the Frequency of Say on Pay Columbia Business School Research Paper No 13-17
Ferri F and T Sandino 2009 The impact of shareholder activism on financial reporting and
compensation The case of employee stock options expensing The Accounting Review 84433ndash
466
Ferri F and J Weber 2009 AFSCME vs Mozilohellipand Say on Pay for All (A) Harvard
Business School Case 109-009
Gompers P J Ishii and A Metrick 2003 Corporate Governance and Equity Prices Quarterly
Journal of Economics 118107-155
Gordon J 2009 Say on Paylsquo Cautionary Notes on the UK Experience and the Case for
Shareholders Opt-In Harvard Journal on Legislation 46323-368
Iliev P and S Vitanova 2013 The effect of the Say on Pay in the US Pennsylvania State
University Working Paper
Kaplan S 2007 Testimony of Steven N Kaplan on Empowering Shareholders on Executive
Compensation and HR 1257 the Shareholder Vote on Executive Compensation Act before the
Committee on Financial Services United States House of Representatives March 8
Larcker D A McCall and G Ormazabal 2013 The Economic Consequences of Proxy
Advisor Say-on-Pay Voting Policies Working Paper Stanford University
Larcker D G Ormazabal and D Taylor 2011 The market reaction to corporate governance
regulation Journal of Financial Economics 101431ndash448
Levit D and N Malenko 2011 Non-binding voting for shareholder proposals Journal of
Finance 661579ndash1614
Mishel L and N Sabadish 2012 How executive compensation and financial-sector pay have
fueled income inequality Issue Brief 331 Economic Policy Institute
Murphy K 2012 The Politics of Pay A Legislative History of Executive Compensation in
Jennifer Hill and Randall Thomas eds Research Handbook on Executive Pay Edward Elgar
Publishers
New York Times 2007 House votes to give investors say on executive pay April 21
Norris F 2004 Corporate democracy and the power to embarrass New York Times March 4
Ozbas O and J Matsusaka2013 Managerial Accommodation Proxy Access and the Cost of
Shareholder Empowerment University of Southern California CLEO Research Paper Series No
C12-1
Thomas R and J Cotter 2007 Shareholder proposals in the new millennium Shareholder
support board response and market reaction Journal of Corporate Finance 13 368ndash391
Thomas R A Palmiter and J Cotter 2012 Dodd-Franks Say on Pay Will it Lead to a Greater
Role for Shareholders in Corporate Governance Cornell Law Review 971213-1266
Thomas R and C Van der Elst 2013 The International Scope of Say on Pay ECGI Law
Working Paper No227
Ferri F and J Weber 2009 AFSCME vs Mozilohellipand Say on Pay for All (A) Harvard
Business School Case 109-009
Gompers P J Ishii and A Metrick 2003 Corporate Governance and Equity Prices Quarterly
Journal of Economics 118107-155
Gordon J 2009 Say on Paylsquo Cautionary Notes on the UK Experience and the Case for
Shareholders Opt-In Harvard Journal on Legislation 46323-368
Iliev P and S Vitanova 2013 The effect of the Say on Pay in the US Pennsylvania State
University Working Paper
Kaplan S 2007 Testimony of Steven N Kaplan on Empowering Shareholders on Executive
Compensation and HR 1257 the Shareholder Vote on Executive Compensation Act before the
Committee on Financial Services United States House of Representatives March 8
Larcker D A McCall and G Ormazabal 2013 The Economic Consequences of Proxy
Advisor Say-on-Pay Voting Policies Working Paper Stanford University
Larcker D G Ormazabal and D Taylor 2011 The market reaction to corporate governance
regulation Journal of Financial Economics 101431ndash448
Levit D and N Malenko 2011 Non-binding voting for shareholder proposals Journal of
Finance 661579ndash1614
Mishel L and N Sabadish 2012 How executive compensation and financial-sector pay have
fueled income inequality Issue Brief 331 Economic Policy Institute
Murphy K 2012 The Politics of Pay A Legislative History of Executive Compensation in
Jennifer Hill and Randall Thomas eds Research Handbook on Executive Pay Edward Elgar
Publishers
New York Times 2007 House votes to give investors say on executive pay April 21
Norris F 2004 Corporate democracy and the power to embarrass New York Times March 4
Ozbas O and J Matsusaka2013 Managerial Accommodation Proxy Access and the Cost of
Shareholder Empowerment University of Southern California CLEO Research Paper Series No
C12-1
Thomas R and J Cotter 2007 Shareholder proposals in the new millennium Shareholder
support board response and market reaction Journal of Corporate Finance 13 368ndash391
Thomas R A Palmiter and J Cotter 2012 Dodd-Franks Say on Pay Will it Lead to a Greater
Role for Shareholders in Corporate Governance Cornell Law Review 971213-1266
Thomas R and C Van der Elst 2013 The International Scope of Say on Pay ECGI Law
Working Paper No227
New York Times 2007 House votes to give investors say on executive pay April 21
Norris F 2004 Corporate democracy and the power to embarrass New York Times March 4
Ozbas O and J Matsusaka2013 Managerial Accommodation Proxy Access and the Cost of
Shareholder Empowerment University of Southern California CLEO Research Paper Series No
C12-1
Thomas R and J Cotter 2007 Shareholder proposals in the new millennium Shareholder
support board response and market reaction Journal of Corporate Finance 13 368ndash391
Thomas R A Palmiter and J Cotter 2012 Dodd-Franks Say on Pay Will it Lead to a Greater
Role for Shareholders in Corporate Governance Cornell Law Review 971213-1266
Thomas R and C Van der Elst 2013 The International Scope of Say on Pay ECGI Law
Working Paper No227